# **Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties

Elia Dal Corso

**STUDY OF THE AINU LANGUAGE**

—

DAL CORSO

Materials and Methods of Analysis for the Study of the Ainu Language

### Ca' Foscari Japanese Studies Linguistics and Language Education

Collana diretta da | A series directed by Luisa Bienati, Paolo Calvetti, Toshio Miyake, Massimo Raveri 18 | 3

### Ca' Foscari Japanese Studies Linguistics and Language Education

#### **General scientific editors**

History and Society sub-series: Toshio Miyake (Università Ca' Foscari Venezia, Italia) Arts and Literature sub-series: Luisa Bienati (Università Ca' Foscari Venezia, Italia) Linguistics and Language Education sub-series: Paolo Calvetti (Università Ca' Foscari Venezia, Italia Religion and Thought sub-series: Massimo Raveri (Università Ca' Foscari Venezia, Italia)

#### **Scientific boards**

History and Society sub-series: Nicola Bassoni (Università degli Studi di Genova, Italia) Sonia Favi (The University of Manchester, UK) Kimio Ito¯ (Kyoto University, Japan) Toshio Miyake (Università Ca' Foscari Venezia, Italia) Andrea Revelant (Università Ca' Foscari Venezia, Italia) Steffi Richter (Universität Leipzig, Deutschland) Sven Saaler (Sophia University, Tokyo, Japan) Hirofumi Utsumi (Otemon Gakuin University, Japan) Marco Zappa (Università Ca' Foscari Venezia, Italia) Arts and Literature sub-series: Tomoko Aoyama (The University of Queensland, Brisbane, Australia) Jaqueline Berndt (Kyoto Seika University, Japan) Luisa Bienati (Università Ca' Foscari Venezia, Italia) Katja Centonze (Università Ca' Foscari Venezia, Italia) Caterina Mazza (Università Ca' Foscari Venezia, Italia) Daniela Moro (Università Ca' Foscari Venezia, Italia) Carolina Negri (Università Ca' Foscari Venezia, Italia) Andreas Regelsberger (Universität Trier, Deutschland) Bonaventura Ruperti (Università Ca' Foscari Venezia, Italia) Silvia Vesco (Università Ca' Foscari Venezia, Italia) Pierantonio Zanotti (Università Ca' Foscari Venezia, Italia)

Linguistics and Language Education sub-series: Patrick Heinrich (Università Ca' Foscari Venezia, Italia) Hideo Hosokawa (Waseda University, Tokyo, Japan) Kikuo Maekawa (National Institute for Japanese Language and Linguistics, Tokyo, Japan) Marcella Mariotti (Università Ca' Foscari Venezia, Italia) Giuseppe Pappalardo (Università Ca' Foscari Venezia, Italia) Barbara Pizziconi (SOAS, University of London, UK) Aldo Tollini (Università Ca' Foscari Venezia, Italia)

Religion and Thought sub-series: Giovanni Bulian (Università Ca' Foscari Venezia, Italia) Andrea De Antoni (Ritsumeikan University, Kyoto, Japan) Federico Marcon (Princeton University, USA) Tatsuma Padoan (University College Cork, Ireland) Silvia Rivadossi (Università Ca' Foscari Venezia, Italia)

#### **Executive coordinator**

Marcella Mariotti (Università Ca' Foscari Venezia, Italia)

#### **Head office**

Dipartimento di Studi sull'Asia e sull'Africa Mediterranea Università Ca' Foscari Venezia Palazzo Vendramin dei Carmini Dorsoduro 3462 30123 Venezia

e-ISSN 2724-2285 ISSN 2724-1203

URL https://edizionicafoscari.unive.it/en/edizioni4/collane/ca-foscari-japanese-studies/

# Materials and Methods of Analysis for the Study of the Ainu Language

Southern Hokkaidō and Sakhalin Varieties

Elia Dal Corso

Venezia **Edizioni Ca' Foscari** - Venice University Press 2022

Materials and Methods of Analysis for the Study of the Ainu Language. Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

© 2022 Elia Dal Corso for the text © 2022 Edizioni Ca' Foscari for the present edition

### cb

Quest'opera è distribuita con Licenza Creative Commons Attribuzione 4.0 Internazionale This work is licensed under a Creative Commons Attribution 4.0 International License

Any part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means without permission provided that the source is fully credited.

Qualunque parte di questa pubblicazione può essere riprodotta, memorizzata in un sistema di recupero dati o trasmessa in qualsiasi forma o con qualsiasi mezzo, elettronico o meccanico, senza autorizzazione, a condizione che se ne citi la fonte.

Scientific certification of the works published by Edizioni Ca' Foscari: the essay published has received a favourable evaluation by subject-matter experts, through a double blind peer review process under the responsibility of the Scientific Committee of the series. The evaluations were conducted in adherence to the scientific and editorial criteria established by Edizioni Ca' Foscari, using a dedicated platform.

Certificazione scientifica delle Opere pubblicate da Edizioni Ca' Foscari: il saggio pubblicato ha ottenuto il parere favorevole da parte di valutatori esperti della materia, attraverso un processo di revisione doppia anonima, sotto la responsabilità del Comitato scientifico della collana. La valutazione è stata condotta in aderenza ai criteri scientifici ed editoriali di Edizioni Ca' Foscari, ricorrendo all'utilizzo di apposita piattaforma.

Edizioni Ca' Foscari Fondazione Università Ca' Foscari | Dorsoduro 3246 | 30123 Venezia http://edizionicafoscari.unive.it | ecf@unive.it

1st edition April 2022 ISBN 978-88-6969-585-8 [ebook] ISBN 978-88-6969-604-6 [print]

Materials and Methods of Analysis for the Study of the Ainu Language. Southern Hokkaidō and Sakhalin Varieties / Elia Dal Corso — 1. ed. — Venice: Edizioni Ca' Foscari, 2022. xii + 164 p.; 23 cm. — (Ca' Foscari Japanese Studies; 18, 3). — ISBN 978-88-6969-604-6.

URL https://edizionicafoscari.unive.it/it/edizioni/libri/978-88-6969-585-8/ DOI http://doi.org/10.30687/978-88-6969-585-8

#### **Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties

Elia Dal Corso

### Abstract

This volume is intended to be a practical manual to learn the basics of the Ainu language, in its varieties of Southern Hokkaidō and Sakhalin. Each lesson presents one specific topic that is investigated taking into account both varieties of the language. Three kinds of activities guide the student throughout each lesson to inductively make generalisations on the language that can be supported by linguistic evidence and, possibly, to revise the information they were given as a start of their analysis. In order to do this, the student learns to observe the data, recognize recurring patterns and exceptions, and formulate a description of the language behaviours illustrated by the given examples. At the end of each lesson the student actively gains knowledge of the Ainu language by producing themselves a set of descriptive rules. The course aims at giving the student some basic analytical tools to approach and analyse Ainu language sources. Furthermore, the student will gain knowledge about the genealogical relationship of the Ainu language and about important sociolinguistic issues relative to its past history and present status of vitality.

**Keywords** Ainu. Language teaching. Linguistics. Endangered languages. Language analysis.

**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

### Acknowledgments

I would like to thank the two anonymous reviewers for their insightful comments both on the general content and on the Ainu examples specifically. My gratitude also goes to Patrick Heinrich, Giuseppe Pappalardo, and Paolo Calvetti for their support of the project that has led to the publication of this volume.

**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

### Table of Contents



### Materials and Methods of Analysis for the Study of the Ainu Language

Southern Hokkaidō and Sakhalin Varieties

**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

### <sup>1</sup> Introductory Lesson: Generalities of the Ainu Language

**Summary** 1.1 The Ainu Language. – 1.2 The Rules of a Language. – 1.3 Linguistic Typology.

#### 1.1 The Ainu Language

Ainu is a language of Far East Asia historically spoken in the territories of today's Russia and Japan that face the Southern Okhotsk Sea. Specifically, Ainu was spoken all through the Japanese island of Hokkaidō (also referred to as 'Ezo/Yezo' in old sources), in the southern part of Sakhalin island (called 'Karafuto' by Japanese people), in the Kuril Islands chain (called 'Chishima' by Japanese people), and in the southernmost tip of the Kamchatka peninsula. Although a number of hypotheses about Ainu's relation to other languages have been proposed (Dougherty 2019, 100), the language is presently classified as an isolate (see Lesson 18). There exist three main varieties of the language, that take their names from the regions where the language was spoken: Hokkaidō Ainu, Sakhalin Ainu, and Kuril Ainu. Both Kuril Ainu and Sakhalin Ainu are extinct, and Hokkaidō Ainu is classified as critically endangered by UNESCO.

> **Ca' Foscari Japanese Studies 18 Linguistics and Language Education 3 DOI 10.30687/978-88-6969-585-8/001**

#### **Dal Corso 1 • Introductory Lesson: Generalities of the Ainu Language**

### Ainu and its neighbour languages

#### 1.2 The Rules of a Language

For any language we can have prescriptive rules and descriptive rules. How do they differ?


In this course we will try to produce descriptive rules from the Ainu data we analyse.

#### **Exercise 1**

Are the following statements prescriptive or descriptive?


#### 1.3 Linguistic Typology

Typology is a discipline of linguistics that studies the variation among languages, measuring its span, differences, and similarities. Typologists create 'groups' of linguistic features that bring languages together and that can be employed to make predictions about those languages.

There exist different types of typology, that refer to different areas of linguistics and are thus defined using different criteria. It follows that languages akin to each other with regards to one typology may differ with regards to another. There are two main typologies:

Morphological typology

• Synthesis index or "how many morphemes are there in one word?"

The answer to this question delineates a continuum at whose extremes we find isolating (or analytic) languages, with one or a few morphemes per word, and polysynthetic languages, with many morphemes per word.


Central Yup'ik (Eskimo-Aleut, USA) is an example of a polysynthetic language.

*Tuntu-ssur-qatar-ni-ksaite-ngqiggte-uq.* reindeer-hunt-fut-say-neg-again-3sg.ind 'He had not yet said again that he was going to hunt reindeer.' (Eliza Orr in Payne 2006, 190)

• Fusivity index or "how many functions/meanings are there *simultaneously* in one morpheme?"

The answer to this question delineates a continuum at whose extremes we find agglutinative languages, with one meaning per morpheme, and fusional languages, with many meanings per morpheme.


Japanese (Japonic, Japan) and Russian (Slavic, Russia) are examples of an agglutinative and a fusional language respectively.

*Mi-saser-u.* see-caus-n.pst 'To make see.' Ja čita-ju knig-u. I read-1S.PRES book-ACC.F.S 'I read a book.'

Syntactic typology

"What is the linear order of the verb and the direct object with respect to each other?"

The answer to this question distinguishes languages in two main groups: VO and OV. This syntactic order of constituents usually has systematic repercussions on the order of words within phrases and clauses within sentences.

VO languages usually have:


OV languages usually have:


#### **Exercise 2**

#### **Ainu (Isolate, Japan) (constructed examples)**

Consider the following clauses and their translation. How much can you understand of the main characteristics of this language? What kind of observations concerning its typology can you make by looking at these examples?


#### **Exercise 3**

#### **Nivkh (Isolate, Russia) (constructed examples)**

Consider the following clauses and their translation. How much can you understand of the main characteristics of this language? What kind of observations concerning its typology can you make by looking at these examples?

Fill in the set with the missing translations. Then, on the basis of the morphosyntactic rules of the language you have deduced from these examples, form one grammatical and one ungrammatical sentence.


Ca' Foscari Japanese Studies 18 | 3 **7** Materials and Methods of Analysis for the Study of the Ainu Language, 3-8

#### **Exercise 4**

#### **Yagua (Peba-Yaguan, Peru) (examples from Payne 2006, 204-5)**

On the basis of the following examples determine whether this language is of the VO or OV type and whether it is of the SV or VS type. Justify your answer. Then list your observations on the order of the other constituents – given your generalisations on syntactic typology, is this constituent order expected or not (or you can't tell)? In the following exercise the asterisk marks ungrammatical examples.


**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

### <sup>2</sup> Phonemic Inventory: The Sounds of the Language

**Summary** 2.1 Background Information and Observation. – 2.2 Research. – 2.3 Analysis and Description.

#### 2.1 Background Information and Observation

Both Hokkaidō Ainu and Sakhalin Ainu (that will be abbreviated as HA and SA respectively from now on) have five vowels, represented in the latin script by *a*, *e*, *i*, *o*, *u*, and twelve consonants, represented in the latin script by *p*, *t*, *k*, *c*, *s*, *r*, *m*, *n*, *w*, *y*, *h*, *'*. The apostrophe is used to signal the glottal stop /ʔ/, a plosive glottal voiceless consonant that consists in an obstruction of the airflow in the glottis and produces a sort of pause like when in English we exclaim *uh-oh*.

This orthography in the latin script represents the following phonemes. I refer the reader to this interactive IPA chart for sound (http://www.ipachart.com/) and to Ladefoged, Maddieson (1996) among others for further discussion:

> **Ca' Foscari Japanese Studies 18 Linguistics and Language Education 3 DOI 10.30687/978-88-6969-585-8/002**


**Table 1** The Ainu orthography and phonemes

The following tables show the place of Ainu vowels and consonants in the IPA charts.

**Table 2** The Ainu vowels

VOWELS

Vowels at right and le of bullets are rounded and unrounded.

**Table 3** The Ainu consonants, labio-velar approximant /w/ and affricate /t ͡ ɕ/ not included


CONSONANTS (PULMONIC)

Where symbols appear in pairs, the one to the right represents a voiced consonant. Shaded areas denote articulations judged impossible.

	- 1. Not all consonants of Ainu are allowed to occur at the end of a word. Precisely, the consonants not allowed word-finally are *c*, *h*, and the glottal stop in HA, and *p*, *t*, *k*, *c*, *r*, and the glottal stop in SA.
	- 2. For plosive consonants there is no voiced-voiceless opposition, so that *p*, *t*, *k* can sometimes be realised respectively as *b* [b], *d* [d], *g* [ɡ] without compromising the comprehensibility of the word and without changing its meaning.
	- 3. The syllable structure is identical in HA and SA. SA might have a long vs. short vowel distinction in open syllables, which would have resulted in two more syllable types (i.e. CVV and VV). The status of long vowels is still debated (see Dal Corso 2021). The table below shows the possible syllable structure in both varieties.
	- 4. Syllable-initially and especially between vowels, the glottal stop may or may not be there depending on the speaker's idiolect, or it may disappear in fast speech.
	- 5. Before /i/ the consonant *s* is realised as [ɕ] (e.g. *sisam* > [ˈɕi.sam]).

#### **Table 4** Syllable structure


#### *Dataset 1 – Plosive consonants*

Consider the following words containing the consonants *p*, *t*, and *k*  (the words are within curly brackets and the consonants under scrutiny are in bold). What is the actual pronunciation of *p*, *t*, and *k* in these words? Is the pronunciation always the same or does it change? If it does change, what does the change depend on? What differences between varieties do you notice? What revisions can you make to the information you were previously given?

#### **Set 1.1** (Hokkaidō Ainu)


#### **Set 1.2** (Sakhalin Ainu)


#### *Dataset 2 – The sibilant consonant* s

Consider the following words containing the consonant *s* (the words are within curly brackets and the consonants under scrutiny are in bold). What is the actual pronunciation of *s* in these words? Is the pronunciation always the same or does it change? If it does change, what does the change depend on? What differences between varieties do you notice? What revisions can you make to the information you were previously given?

#### **Set 2.1** (Hokkaidō Ainu)


#### **Set 2.2** (Sakhalin Ainu)


#### 2.2 Research

Now look at this other dataset and also read the examples from other languages given after it, which you will need for the third and last activity of this lesson.

#### *Dataset 3 – The consonants* r *and* h

Consider the following words containing *r* and *h* (the words are within curly brackets and the consonants under scrutiny are in bold). What is the actual pronunciation of *r* and *h* in these words? Is the pronunciation always the same or does it change? If it does change, what does the change depend on? What differences between varieties do you notice?

#### **Set 3.1** (Hokkaidō Ainu)


#### **Set 3.2** (Sakhalin Ainu)


#### **Examples from Other Languages**

Italian (Romance, Italy)

What does the distribution of the sounds /p/ and /t/ in Italian tell us?


In Italian, the distribution and overall the presence of the sounds /p/ and /t/ in a word cannot be **predicted** on the basis of the other sounds in that word – see the words 'toro' and 'poro' that have, besides /p/ and /t/, the same vowels and the same consonant (i.e. they are **minimal pairs**). The sounds /p/ and /t/ can appear in the same **phonetic environment**, in words with different meanings – they are therefore said to be in **contrastive distribution**. When two sounds are in contrastive distribution and their alternation causes a change in a word's meaning then these sounds represent **(separate) phonemes** of that language.


As exemplified in this limited data set, in Hupa there are no minimal pairs that are differentiated by the vowels [ɪ] and [u]. If we look at the sounds occurring before *and* after [ɪ] and [u] we do not find any identical phonetic environment where they appear – they are not in contrastive distribution, but rather in **complementary distribution**. This means that in some phonetic environments we will find one sound and in others the other sound – the distribution of [ɪ] and [u] is **predictable**. From the examples above we see that [u] only occurs before a labial-velar sounds (either [ʍ] or [w]), while [ɪ] never appears in this environment. We can therefore say that [ɪ] and [u] are **allophones** (different phonetic realisations) of the same phoneme. One or the other realisation depends on the environment where the sound is found. Since [ɪ] appears in the **most varied environments** (i.e. in between all kinds of consonants) but [u] appears only before a labial-velar consonant, we can conclude that /ɪ/ is the **underlying phoneme** (i.e. the 'original' sound) and [u] is the realisation it takes when in presence of a labial-velar consonant. Formally, we can write this rule we just discovered as: ɪ > |u| / \_ʍ; \_w OR ɪ > |u| / \_C[labial-velar] ɪ > | ɪ | / [elsewhere]

#### Korean (Koreanic, Korea)


In Korean, stops at the end of a word (and generally in coda position within a syllable) are realised as unreleased. We signal the lack of audible sound with ' ̚ ' in IPA.


#### Gujarati (Indo-Aryan, India) (examples from Esposito, Dowla Khan 2012)

Tausug (Austronesian, Philippines)


In Gujarati, vowels can be pronounced with less strength, almost as breathing or whispering, This happens under the influence of particular syllable structure, tone, and other factors. A **breathy vowel** is signaled in IPA with ' ͈ '.

#### 2.3 Analysis and Description

Once you are finished with your analysis of the data, describe in no less than 300 words the phonetic changes that characterise *r* and *h*  in Ainu citing the necessary examples from the analysed data (optionally also describe the phonetic changes occurring for the consonants taken into account in datasets 1 and 2). Try to comment on the following in your description.


**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

# <sup>3</sup> Personal Affixes, Reciprocal and Reflexive

**Summary** 3.1 Background Information and Observation. – 3.2 Research. – 3.3 Analysis and Description.

#### 3.1 Background Information and Observation

HA makes a distinction among first, second, third, and fourth person, while SA only distinguishes first, second, and third person. The term 'fourth person' is actually an improper term since it could be easily misunderstood as referring to the obviative person in a directinverse alignment (see e.g. Jacques, Antonov 2014). Tamura (2000) prefers to discuss it as 'indefinite person'. However, since referencing to an indefinite agent is only one of the functions of fourth person, this latter term has found its place in Ainu studies and it continues to be widely employed.

The so-called fourth person of HA is used to mark a first person plural inclusive (that is, a first person plural 'we' that includes the speaker and the interlocutor) (1), as opposed to the first person plural exclusive (that is, a first person plural 'we' that includes the speaker, some other referent(s) but not the interlocutor). The fourth person is otherwise used to mark an indefinite agent (2), a second person honorific (3), or a first person in direct speech (4).


**Ca' Foscari Japanese Studies 18 Linguistics and Language Education 3 DOI 10.30687/978-88-6969-585-8/003**


In SA there is no separate fourth person. In this Ainu variety the personal affixes *an-/-an*, cognates of the HA ones, are treated as the default forms for first person plural (Murasaki 1979). Nevertheless, the affixes overlap functionally in the two varieties to some extent, enough that the meanings illustrated in examples (2) and (4) are also attested for the SA *an-/-an* (5)-(6). For this reason, it makes sense to acknowledge the presence of a polysemic 'fourth person' in SA as well (Dal Corso 2021). Differently from HA, the fourth person forms in SA have become the only way to express first person plural.

There are of course some differences with the HA fourth person. Since there is no separate marking for fourth person and first person plural, a formal distinction between first person plural inclusive and exclusive is absent. Moreover, fourth person markers in SA are not used to cross-reference an honorific second person.


'She thought: "How in the world [can] **I** do to help or assist it?"' (Dal Corso 2021)

Person is marked on Ainu verbs via affixes (both prefixes and suffixes). In both HA and SA third person singular is signalled by the lack of affixes. While for HA this is also true for third person plural, in SA this latter can sometimes be marked overtly. **Personal affixes are only used for the subject and the object (or indirect object) of a verb and a maximum of two affixes can be attached to a verb**. The following tables show all personal agreement markers (subject referent) of the verbal paradigm on an intransitive and a transitive verb.


**Table 1** Personal agreement affixes of HA and SA with the intransitive verb *mina*/*miina* 'laugh'

**Table 2** Personal agreement affixes of HA and SA with the transitive verb *nukar*/*nukara* 'see, look', with implied third person singular object


*Consider this additional information…*


#### *Dataset 1 – Transitive agreement*

Consider the following example sentences and the relative translations (sentences marked with \* are ungrammatical). Can you determine the full transitive agreement paradigm of Ainu verbs? Try to fill in the tables below. Can you determine what is the correct version of sentences marked as ungrammatical? What peculiarities do you notice? What generalisations on the syntactic typology of the language can you make? What revisions/additions to the previously given information can you propose?



Transitive agreement paradigm (HA)


Transitive agreement paradigm (SA)


Once you are finished, compare your results with the tables in the appendix.

#### 3.2 Research

Now look at this other dataset and also read the examples from other languages given after it, which you will need for the third and last activity of this lesson.

#### *Dataset 2 – Phonological changes*

Consider the verb forms that appear in the following short sentences. All verb forms are given in IPA (note that: indicates a long vowel/double consonant) and they feature some of the intransitive and transitive personal agreement affixes, and the reflexive and reciprocal prefixes *yay-* [jaj-] and *u-* [u-] that you have just analysed. What changes do you notice in their phonological realisation or in the realisation of the verb they are attached to?



41. t ͡ɕeh et ͡ɕiʔuɸ You all take the fish. 42.ten:ehpo kuo:nohka I lull the baby. 43. ajnu anʔa:ɾe I make the man sit.

#### **Examples from Other Languages…**

Ancient Greek (Hellenic, Greece)

γράφειν 'to write': perfect tense \*γεγραφ- /ɣeɣraɸ/


In Ancient Greek the perfect tense of the verb γράφειν is formed from the stem \*γεγραφ- [ɣeɣraɸ], that is made up of two **bound morphemes**: the verbal root γραφ- and the prefix Cε- (realised as γε- in this instance). When personal endings are added, the final sound of the verbal stem and the first sound of the personal suffix may be difficult to pronounce in sequence. Therefore, one of the two sounds is accomodated to ease pronunciation – one sound undergoes **assimilation** to the other. If the change concerns the second sound in the sequence, assimilation is said to be **progressive**. If the change concerns the first sound in the sequence, assimilation is said to be **regressive**. The case above shows two instances of regressive assimilation in the forms ɣeɣrammai, ɣeɣrammeθa and ɣeɣraptai, where the voiceless bilabial fricative /ɸ/ assimilates to the bilabial nasal /m/ and to the voiceless dental-alveolar plosive /t/ respectively, acquiring the same manner of articulation thus becoming itself the bilabial nasal [m] in the first case, or the voiceless bilabial plosive [p] in the second case. In all other phonetic environments /ɸ/ remains unchanged since we do not encounter difficulties in pronunciation. We can formalise this assimilation as follows: /ɸ/ > [m] / \_m /ɸ/ > [p] / \_t /ɸ/ > [ɸ] / elsewhere The phonological change that concerns /ɸ/ in the form ɣeɣrapsai is, on the contrary, a case of dissimilation (see below).

#### Sanskrit (Indo-Aryan, India)


Declension of the noun /ɡo:dʰuk/ 'cow-milker'.

Throughout its declension, the Sanskrit noun ɡo:dʰuk 'cow-milker' features two phonological processes. The first process is assimilation (i.e. /k/ > [ɡ] / \_Cvoiced) and the second is **dissimilation** which is the opposite of assimilation. With dissimilation a sound becomes less similar to another to ease pronunciation. In the case at hand the aspiration imposed on the voiced dental-alveolar plosive /dʰ/ is lost whenever the voiceless glottal fricative /h/ appears in the noun ending. Like assimilation, dissimilation can be progressive or regressive. We formalise dissimilation in Sanskrit as follows: /dʰ/ > [d] / \_uh

#### Hanunó'o (Austronesian, Philippines) (data from Olson, Schultz 2002 in Payne 2006, 87)


In Hanunó'o numeral adverbs are formed with the addition of the prefix *ka-* to the cardinal number form. However, the first vowel of the cardinal number form disappears if it is *u*. This phonological process is called **elision** and again it serves to ease pronunciation. In addition to elision, cardinal numbers where *u* is preceded by the glottal stop also show **metathesis** – the glottal stop and the consonant following the elided *u* are inverted. We can formalise elision in adverbial numerals as follows: /u/ > Ø / C\_C We can formalise elision followed by metathesis as:

/u/ > Ø / ʔ \_ C; ʔCi > Ci ʔ

#### Ancient Greek (Hellenic, Greece)

#### Declension of the noun ἀνήρ 'man' /anɛr/ (root \*ανρ- /anr/)


The declension of the noun ἀνήρ 'man' in Ancient Greek starts from the root \*ανρ- / anr/ that contains a consonant cluster that is difficult to pronounce (i.e. /nr/). In order to ease pronunciation, an epenthetic sound is added – this process is called **insertion** (or **epenthesis**) and it is the opposite of elision.

We formalise insertion as it appears in the case at hand as follows: Ø > [d] / n\_r

#### English (Germanic, UK) (data from Payne 2006, 76)


Morphemes can take different realisations depending on the **phonological environment** they are in. Sometimes, the phonological processes a morpheme undergoes (among which the ones above) can affect its realisation to the point where it is difficult to recognise the separate forms actually as the same morpheme. Usually semantics helps us in this case, giving us a hint of the same source for the separate forms we see. The prefixes [ɪn-], [ɪm-] and [ɪŋ-] in English are one such case, for which we can assume a same origin given the systematic meaning they contribute to the word they attach to (i.e. a negative meaning). To make linguistic analysis and description easier, morphemes are presented in one 'representative' form which may change due to phonological processes – this form is the **underlying form**. The underlying form should be the form that occurs in the largest number of environments and the one that is most difficult to derive by a rule – these two precautions save us a lot of time in writing derivation rules! In the case at hand, [ɪn-] appears before æ, d, t, s, v, [ɪm-] appears before p, m, b, and [ɪŋ-] appears before k, ɡ. The most varied environment is the one of [ɪn-] that we take as our underlying form, while the other two realisations are **alloforms** of this latter. We can formalise the derivation rules as follows:

/ɪn-/ > [ɪm-] / \_C[bilabial] /ɪn-/ > [ɪŋ -] / \_C[velar] /ɪn-/ > [ɪn-] / elsewhere See Payne 2006, 70-3 for further explanation.

#### 3.3 Analysis and Description

Once you finish your analysis of the data, describe, in no less than 300 words, the phonological processes that characterise personal affixes of HA and SA. Provide the right name for them and, optionally, write a formal rule to describe them.


**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

### <sup>4</sup> Inalienable Possessive Forms

**Summary** 4.1 Background Information and Observation. – 4.2 Research. – 4.3 Analysis and Description.

#### 4.1 Background Information and Observation

HA makes a distinction among first, second, third, and fourth person, while SA only distinguishes first, second, and third person. The term 'fourth person' is actually an improper term since it could be easily misunderstood as referring to the obviative person in a directinverse alignment (see e.g. Jacques, Antonov 2014). Tamura (2000) prefers to discuss it as 'indefinite person'. However, since referencing to an indefinite agent is only one of the functions of fourth person, this latter term has found its place in Ainu studies and it continues to be widely employed.

The so-called fourth person of HA is used to mark a first person plural inclusive (that is, a first person plural 'we' that includes the speaker and the interlocutor) (1), as opposed to the first person plural exclusive (that is, a first person plural 'we' that includes the speaker, some other referent(s) but not the interlocutor). The fourth person is otherwise used to mark an indefinite agent (2), a second person honorific (3) and a first person in direct speech (4).



(4) *Ta-p 'a-e-\_upaskuma tane ka ki … sekor Haca sekor*  this-thing **4S-**APPL-3SO/tell now even SLV/VO/do COMP Haca COMP *'a-ye 'acapo hawean.* 4S-3SO/say old.man 3SS/speak 'An old man called Haca said: "indeed now **I** tell these facts …"' (Tamura 1984, 22)

In SA there is no separate fourth person. In this Ainu variety the personal affixes *an-/-an*, cognates of the HA ones, are treated as the default forms for first person plural (Murasaki 1979). Nevertheless, the affixes overlap functionally in the two varieties to some extent, enough that the meanings illustrated in examples (2) and (4) are also attested for the SA *an-/-an* (5)-(6). For this reason, it makes sense to acknowledge the presence of a polysemic 'fourth person' in SA as well (Dal Corso 2021). Differently from HA, the fourth person forms in SA have become the only way to express first person plural.

There are of course some differences with the HA fourth person. Since there is no separate marking for fourth person and first person plural, a formal distinction between first person plural inclusive and exclusive is absent. Moreover, fourth person markers in SA are not used to cross-reference an honorific second person.


'She thought: "How in the world [can] **I** do to help or assist it?"' (Dal Corso 2021)

Person is marked on Ainu verbs via affixes (both prefixes and suffixes). In both HA and SA third person singular is signalled by the lack of affixes. While for HA this is also true for third person plural, in SA this latter can sometimes be marked overtly. **Personal affixes are only used for the subject and the object (or indirect object) of a verb and a maximum of two affixes can be attached to a verb**. The following tables show all personal agreement markers (subject referent) of the verbal paradigm on an intransitive and a transitive verb.



**Table 2** Personal agreement affixes of HA and SA with the transitive verb *nukar*/*nukara* 'see, look', with implied third person singular object


#### *Consider this additional information…*

HA distinguishes between alienable and inalienable possession. When property of something can be suspended or removed, possession is said to be alienable. On the contrary, possession is inalienable when the property of something cannot be denied nor suspended and resumed – this kind of possession is common for nouns referring to body parts, bodily excretions, relatives and family, personal utensils, etc.

HA employs two distinct constructions to mark alienable and inalienable possession. Alienable possession is expressed with an analytic construction involving the verb *kor* 'have' on which the possessor is cross-referenced via transitive personal agreement markers in the subject form (7), while inalienable possession is expressed through dedicated morphology in a synthetic construction (8) (see e.g. Tamura 2000, 81-9; Bugaeva 2004, 19-21).


In the synthetic possessive construction the possessor is again crossreferenced via transitive personal agreement markers in the subject form attached on the possessed noun. This latter noun also hosts the possessive suffix, which is *-hu* in (8).

The vowel in the possessive suffix is not fixed, but it is copied from the final vowel of the noun stem the suffix attaches to – e.g. *kampi*  'letter' > *kampihi* 'his/her/their letter', *sapa* 'head' > *sapaha* 'his/her/ their head'. Therefore, we can provisionally treat the possessive suffix as having the form *-hV*. If the noun stem ends in a consonant, the possessive suffix takes the form *-VhV* and the copied vowel is the preconsonant one in the noun stem – e.g. *kisar* 'ear' > *kisaraha* 'his/her/ their ear', *tek* 'hand' > *tekehe* 'his/her/their hand'.

SA only displays the synthetic possessive construction for all nouns – formally there is no distinction between alienable and inalienable possession (Murasaki 1979, 81-4).

#### *Dataset 1 – Transitive agreement*<sup>1</sup>

Consider the following example sentences and the relative translations (sentences marked with \* are ungrammatical). Can you determine the full transitive agreement paradigm of Ainu verbs? Try to fill in the tables below. Can you determine what is the correct version of sentences marked as ungrammatical? What peculiarities do you notice? What generalisations on the syntactic typology of the language can you make? What revisions/additions to the previously given information can you propose?

1 For the reader's convenience, in this lesson I intentionally repeat Dataset 1 and part of the examples from other languages at the end of the activities from Lesson 3, since a proper understanding of the possessive forms' morphology featured in Dataset 2 follows from having completed (or revised) this activity on personal agreement affixes.


**Set 1.2** (Sakhalin Ainu) 1. *Eciinuu.* You all hear us. 2. *Tan aynu etura.* This person accompanies you. 3. *Wen aynu seta koyki.* The bad man beats up the dog. 4. *Ceh ankoyki.* We beat up (= kill) the fish. 5. *Ecikoykiyan.* I beat you all. 6. *\* Eentura.* You accompany me. 7. *Annuuhci.* We hear them. 8. *Ekasi ucaskuma nuu.* The old man hears the tale. 9. *Ennuu.* You hear me. 10. *Ukoykian.* We beat up (= strike) ourselves. 11. *\* Kuenuu.* I hear you. 12. *Ahci ecitura.* You all accompany the old woman. 13. *Esetaha ekoykihci.* You beat up your dogs. 14. *Ennuuhci.* Someone hears me. 15. *\* Kuenkoyki.* I beat up (= strike) myself. 16. *Ecikoykihci.* They beat you all. 17. *Enturayan.* You all accompany me. 18. *Ahciutah seta koyki.* The old women beat up the dog. 19. *Annuu.* We hear them. 20. *Inuu.* You hear us. 21. *Ahci ka ekasi ka eciturahci.* You all accompany the old woman and the old man. 22.*Ahci nuuhci.* The old woman hears them. 23.*Ecikoyki.* I beat you up. 24. *\* Anikoyki.* We beat up ourselves. 25. *Iturahci.* They accompany us. 26. *Inuuyan.* You all hear us. 27. *Ahci kunuu.* I hear the old woman. 28. *Anetura.* We accompany you. 29. *Ahci ka ekasi ka seta koykihci.* The old woman and the old man beat up the dog. 30. *Ahciutah kuturahci.* I accompany the old women. 31. *Aynu inuu.* The man hears us. 32. *Ekoykihci.* They beat you up. 33. *Ahciutah ekasi ka seta ka nuuhci.* The old women hear an old man and a dog. 34. *Ekasi ecitura.* The old man accompanies you all. 35. *\* Ahci nuuhci.* The old woman hears someone. 36. *Ecinuuyan.* We hear you all. 37. *Ceh ekoyki.* You beat up (= kill) the fish. 38. *Aynu 'enkoyki.* The man beats me up.

Transitive agreement paradigm (HA)


Transitive agreement paradigm (SA)


Once you are finished, compare your results with the tables in the appendix.

#### 4.2 Research

Now look at this other dataset and also read the examples from other languages given after it, which you will need for the third and last activity of this lesson.

#### *Dataset 2 – Phonological changes*

Consider the examples given below that show some nouns in the nonpossessive and in the possessive forms. All noun forms are given in IPA (note that: indicates a long vowel/double consonant; ungrammatical examples are marked with \*). What changes do you notice in the phonological realisation of the possessive suffix *-(V)hV* and in that of the personal agreement prefixes cross-referencing the possessor? Do you notice any behaviors at odds with the information you were previously given?

**Set 2.1** (Hokkaidō Ainu) 1. topa 'group' topaha 'his/her/their group' 2. set̚ 'nest' kuset ͡ɕiɕi 'my nest' 3. ɕike 'luggage' aɕikehe 'our (inc.) luggage' 4. kotpaɾa̤ 'chest' ekotpaɾoho 'your chest' 5. nan 'face' kunanuɸu 'my face' 6. kewtum 'feeling' kewtumuɸu 'his/her/their feeling' 7. ʔiɾi̤ wak̚ 'sibling' kuiɾi̤ wakiɕi 'my sibling' 8. kew 'body' et ͡ɕikewehe 'you all's body' 9. etu 'nose' eʔetuɸu 'your nose' 10. kujʔoj 'bladder' kujʔojehe 'his/her/their bladder' 11.ɸut ͡ɕi 'grandmother' \* et ͡ɕiɸut ͡ɕiɕi 'you all's grandmother 12.otop̚ 'hair' t ͡ɕotopiɕi 'our (exc.) hair' 13. ɕik̚ 'eye' kuɕikiɕi 'my eye' 14. emko 'half' emkoho 'his/her/their half' 15.ɾu 'path' t ͡ɕiɾuwehe 'our (exc.) path' 16. etu 'nose' e:tuɸu 'your nose' 17. osoɾo̤ 'buttocks' kosoɾoho 'my buttocks' 18. mat̚ 'wife' mat ͡ɕiɕi 'his/her/their wife' 19. hapo 'mother' koɾo̤ hapo 'his/her/their mother' 20. kotpaɾa̤ 'chest' ekotpaɾo 'your chest' 21.topa 'group' topa 'his/her/their group' 22.uni 'house' kuniɕi 'my house' 23.uni 'house' et ͡ɕiʔuniɕi 'you all's house' 24. iɾi̤ wak̚ 'younger brother' t ͡ɕiɾi̤ wakiɕi 'our (exc.) younger brother' 25. kisaɾa̤ 'ear' ekisaɾaha 'your ear' 26. ɕike 'luggage' kuɕike 'my luggage' 27. hon 'belly' ahoniɕi 'our (inc.) belly' 28.haw 'voice' kuhawehe 'my voice' 29. kuɾṳ 'shadow' ekuɾiɕi 'your shadow' 30.ɸut ͡ɕi 'grandmother' et ͡ɕikoɾo̤ ɸut ͡ɕi 'you all's grandmother' 31. iɾo 'colour' kujɾoho 'my colour' 32. hoku 'husband' ehokuɸu 'your husband' 33.uni 'house' et ͡ɕiuniɕi 'you all's house' 34.hapo 'mother' \* hapoho 'his/her/their mother' 35. kujʔoj 'bladder' kujʔoje 'his/her/their bladder'

#### **Set 2.2** (Sakhalin Ainu)


Ca' Foscari Japanese Studies 18 | 3 **37** Materials and Methods of Analysis for the Study of the Ainu Language, 29-42


#### **Examples from Other Languages…**

#### Ancient Greek (Hellenic, Greece)

γράφειν 'to write': perfect tense \*γεγραφ- /ɣeɣraɸ/


In Ancient Greek the perfect tense of the verb γράφειν is formed from the stem \*γεγραφ- [ɣeɣraɸ], that is made up of two **bound morphemes**: the verbal root γραφ- and the prefix Cε- (realised as γε- in this instance). When personal endings are added, the final sound of the verbal stem and the first sound of the personal suffix may be difficult to pronounce in sequence. Therefore, one of the two sounds is accomodated to ease pronunciation – one sound undergoes **assimilation** to the other. If the change concerns the second sound in the sequence, assimilation is said to be **progressive**. If the change concerns the first sound in the sequence, assimilation is said to be **regressive**. The case above shows two instances of regressive assimilation in the forms ɣeɣrammai, ɣeɣrammeθa and ɣeɣraptai, where the voiceless bilabial fricative /ɸ/ assimilates to the bilabial nasal /m/ and to the voiceless dental-alveolar plosive /t/ respectively, acquiring the same manner of articulation thus becoming itself the bilabial nasal [m] in the first case, or the voiceless bilabial plosive [p] in the second case. In all other phonetic environments /ɸ/ remains unchanged since we do not encounter difficulties in pronunciation. We can formalise this assimilation as follows: /ɸ/ > [m] / \_m /ɸ/ > [p] / \_t /ɸ/ > [ɸ] / elsewhere The phonological change that concerns /ɸ/ in the form ɣeɣrapsai is, on the contrary,

a case of dissimilation (see below).

#### Sanskrit (Indo-Aryan, India)

Declension of the noun /ɡo:dʰuk/ 'cow-milker'.


Throughout its declension, the Sanskrit noun ɡo:dʰuk 'cow-milker' features two phonological processes. The first process is assimilation (i.e. /k/ > [ɡ] / \_Cvoiced) and the second is **dissimilation** which is the opposite of assimilation. With dissimilation a sound becomes less similar to another to ease pronunciation. In the case at hand the aspiration imposed on the voiced dental-alveolar plosive /dʰ/ is lost whenever the voiceless glottal fricative /h/ appears in the noun ending. Like assimilation, dissimilation can be progressive or regressive. We formalise dissimilation in Sanskrit as follows: /dʰ/ > [d] / \_uh

#### Hanunó'o (Austronesian, Philippines) (data from Olson, Schultz 2002 in Payne 2006, 87)


In Hanunó'o numeral adverbs are formed with the addition of the prefix *ka-* to the cardinal number form. However, the first vowel of the cardinal number form disappears if it is *u*. This phonological process is called **elision** and again it serves to ease pronunciation. In addition to elision, cardinal numbers where *u* is preceded by the glottal stop also show **metathesis** – the glottal stop and the consonant following the elided *u* are inverted. We can formalise elision in adverbial numerals as follows: /u/ > Ø / C\_C

We can formalise elision followed by metathesis as:

/u/ > Ø / ʔ \_ C; ʔCi > Ci ʔ

If elision concerns the final segment of a word, it is otherwise called **truncation** or **apocope**.

#### Niimiipuutímt (Plateau Penutian, USA) (data from Aoki 1970)


Vowels in Niimiipuutímt are divided into two classes – dominant and recessive. This subdivision serves to account for the phonological changes we witness when morphemes containing certain vowels attach to other morphemes. In this language /o/ and /ɑ/ are dominant vowels and /æ/ and /u~ɯ/ are recessive vowels. The vowel /i/ belongs to both groups. When two or more morphemes combine, if any of these morphemes contain a dominant vowel then all recessive vowels present in all the morphemes of that word need to change to their dominant counterpart (see the first singular possessive /næʔ-/ that changes into /nɑʔ-/ under the influence of the dominant vowel /o/ in the noun root /tot/). The reverse does not happen. This phenomenon is called **vowel harmony** and it is a kind of long-distance assimilation, by which two or more non-contiguous sounds assimilate. Since /i/ belongs to both vowel groups it may or may not trigger vowel harmony. We can formalise vowel harmony for the first singular possessive prefix of Niimiipuutímt as follows: V > V[front] / \_CV[front]



Morphemes can take different realisations depending on the **phonological environment** they are in. Sometimes, the phonological processes a morpheme undergoes (among which the ones above) can affect its realisation to the point where it is difficult to recognise the separate forms actually as the same morpheme. Usually semantics helps us in this case, giving us a hint to a same source for the separate forms we see. The prefixes [ɪn-], [ɪm-] and [ɪŋ-] in English are one such case, for which we can assume a same origin given the systematic meaning they contribute to the word they attach to (i.e. a negative meaning). To make linguistic analysis and description easier, morphemes are presented in one 'representative' form which may change due to phonological processes – this form is the **underlying form**. The underlying form should be the form that occurs in the largest number of environments and the one that is most difficult to derive by a rule – these two precautions save us a lot of time in writing derivation rules! In the case at hand, [ɪn- ] appears before æ, d, t, s, v, [ɪm-] appears before p, m, b, and [ɪŋ-] appears before k, ɡ. The most varied environment is the one of [ɪn-] that we take as our underlying form, while the other two realisations are **alloforms** of this latter. We can formalise the derivation rules as follows:

/ɪn-/ > [ɪm-] / \_C[bilabial] /ɪn-/ > [ɪŋ -] / \_C[velar] /ɪn-/ > [ɪn-] / elsewhere See Payne 2006, 70-3 for further explanation.

#### 4.3 Analysis and Description

Once you finish your analysis of the data, describe, in no less than 300 words, the phonological processes that characterise personal affixes cross-referencing the possessor in HA and SA and the possessive suffix (if any, highlight problematic cases). Try to provide the right name for these processes and, optionally, write a formal rule to describe them.


**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

# <sup>5</sup> The Ainu Passive Construction

**Summary** 5.1 Background Information and Observation. – 5.2 Research. – 5.3 Analysis and Description.

#### 5.1 Background Information and Observation

Among the uses of the HA fourth person Tamura (2000, 71-2) discusses that of cross-referencing the starting point of the action in a passive construction. A sentence like (1) is regarded as a passive sentence by virtue of the fact that the actual performer of the event is expressed in a noun phrase with *'orowa*, an oblique. The performer of the action is referenced to by the fourth person prefix on the verb. This is different from the impersonal construction (2) where the indefinite agent is still marked via the fourth person on the verb but it is not overtly expressed.


In the active version of sentences like (1) the agent is no longer marked by *'orowa* and it is cross-referenced to via third person null agreement on the verb.

(3) *Hapo en-koyki.* mother 1SO-**3SS/**scold. 'Mother scolds me.'

Murasaki (1979), on the other hand, does not report any analogous passive construction in SA. In this variety of the language only the impersonal construction is reported, where the indefinite agent is cross-referenced to via the third person plural suffix *-hci* (4). This construction may be read as passive depending on context.

(4) *Taa cise-'oh-ta 'i-'ama-hci 'anah pirika kusu ne ya.* that house-place-in 4O-put**-3PS** if 3SS/be.good INTN FP 'It would be good if someone left us at that house [= if we were left at that house]' (Murasaki 1979, 51)

If the third person agent is overtly expressed the construction is no longer impersonal and the passive reading is cancelled.

*Consider this additional information…*

We can trace four important stages in the development of the HA passive (Bugaeva 2011), that resulted in the construction attested in the textual sources:


Consider the following sentences (\* marks ungrammatical examples, ? marks dubious examples). What is the typological order of constituents in the passive construction? What additions and amendments can you make to the information you started from? What syntactic status do the constituents in the construction have (argument/ oblique)? Justify your answer. Is the passive construction of HA and SA analytic or synthetic? Motivate your answer. What differences do you notice between HA and SA?

**Set 1.1** (Hokkaidō Ainu)

	- 'The tree was broken by the wind.' (Bugaeva 2004, 41)

'Her grandmother loves the girl.'


10. *Aiipere katu ka isam.* 'There was no sign [that] someone [would] let me eat [= I would be given food].'

(Tamura 1985, 2)

11. *\* Toan nispaorowa taan kur kusa a ruwe?*

'Has this man been taken across [the river] by that noble man?'

12. *Kamuyorwa ene aiyaynure.*

'I was taken in such consideration by a god.' (Tamura 1985, 46)

13. *Matkaci suciorowa aomap.*

'The girl is loved by her grandmother.' (Bugaeva 2004, 42)

#### **Set 1.2** (Sakhalin Ainu)


#### 5.2 Research

Now look at this other dataset and also read the examples from other languages given after it, which you will need for the third and last activity of this lesson.

#### *Dataset 2 – Syntactic tests*

Consider the following clauses paying particular attention to ungrammatical examples (marked with \*). Only HA examples are provided for this dataset. The passive construction is here used in combination with three other constructions – what are they and how are they expressed morphosyntactically? Why do you think ungrammaticality arises?

#### **Set 2** (Hokkaidō Ainu)


'He carried back a stone.'


'I only had a mother.' (Tamura 1984, 36)

12. *\* Huciorowa aenomap rusuy.*

'I want to be loved by grandmother.'

13. *Siosmakun kuinkar.*

'I looked behind myself.' (Satō 1995, 9)

14. *Horipian kor…*

'While I danced…' (IH 1985, line 287)

15. *Eper* […] *nimakihi ka* […] *amihi ka nispaorowa atuypa.*

'The bear's nails and teeth were cut off by the noble man.' (adapted from Bugaeva 2011, 536)

#### **Examples from Other Languages…**

#### Japanese (Japonic, Japan)

*Sensei-ga gakusei-o yob-u. Gakusei-ga sensei-ni yob-a-rer-u.* teacher-NOM student-ACC call.NPST student-NOM teacher-by call-INFL-PASS-NPST 'The teacher calls the student.' 'The student is called by the teacher.'

The Japanese passive is a synthetic construction since passive voice is expressed on the verb via the suffix -*(ra)rer*. In the (direct) passive construction what was the object in the active sentence is **promoted to subject** and it gets assigned subject morphology (i.e. the NOM postposition *-ga*), while the subject of the active sentence is **demoted to an oblique role** and marked by *-ni*. The passivised verb is now grammatically **intransitive**. These characteristics make the Japanese -*(ra)rer*  construction a prototypical passive.

Nivkh (Isolate, Russia) (data from Nedjalkov, Otaina 2012, 255)

*ətək p'-su-d̹ .* father REF-wash-IND 'Father washes himself.'

*ətək p'-ōla-du-d̹ .* father REF-child-wash-IND 'Father washes his own son.'

*umgu ñivx-ñřə-r p'-ōla-du-d̹ .* woman man-see-CONV.NAR.3SG REF-child-wash-IND 'The woman sees the man and washes \*his / her own son.' (constructed example)

*ətək ōla-aχ p'-su-gu-d̹ .* father child-DAT/ACC REF-wash-CAUS-IND 'Father has his son wash him(= father) / himself(= son).'

The personal/possessive reflexive pronoun *p'-/p'i* in Nivkh can only refer to the subject. This is the rule of **reflexive binding** in Nivkh and in the majority of languages. In the third sentence given here *umgu* 'woman' is the subject of both coordinate clauses (implied for the verb 'wash'), while *ñivx* 'man' and *ōla* 'child' are objects respectively of the first and second clause. Although it could be contextually possible for the the child to be the man's, this reading of the sentence is ruled out by the use of the possessive reflexive *p'-* which can only bind the subject of the sentence (i.e. the woman). In order to express the meaning 'his (= the man's) son' a third person possessive pronoun would be needed. The last sentence shows a causative construction. Causativisation results in the promotion of an originally (more) peripheral argument to a (more) core argument. Here what is marked as the indirect object (with the suffix -aχ) of the causative verb 'make wash' is actually the logical subject of the non-causative verb 'wash'. That is, despite its morphological marking the noun *ōla* also functions as a subject in this case. Therefore reflexive binding becomes ambiguous and the double reading of the reflexive pronoun is possible.

#### Kalaallisut (Inuit, Greenland)

(data from and adapted from Bjørnum 2012, 86-8)

*Qimmi-a sinip-poq.* dog-POSS.ABS.3SG sleep-3SG 'His dog sleeps.'

*Qimmi-ata kiiva-anga.* dog-POSS.ERG.3SG bit-3SG>1SG 'His dog bit me.'

*Piita-p qimmi-a nerukkar-paa.* Peter-ERG dog-POSS.ABS.3SG fed-3SG>3SG 'Peter fed his (= someone else's) dog.'

Kalaallisut (West Greenlandic) has morphological **ergative-absolutive alignment**. Besides being evident in the marking of verbal arguments, the ergative-absolutive alignment also concerns possession that is expressed with suffixes on the possessed noun. These suffixes agree in number and person with the possessor and they are also marked for ergative or absolutive case according to the grammatical function covered by the possessee.


Impersonal constructions are defined as such due to the **lack of reference** to a specific agent. In some languages like Gude a generic noun or pronoun can be used to refer to an **unspecified agent** (e.g. *ənji* '(some) people') and it functions as the subject of the active sentence, which may also have a passive reading. In other languages like Paamese, the expression of the agent (even an unspecified one) via a separate noun phrase is not possible and impersonality is solely expressed by verbal morphology (in Paamese, third person agreement).

#### 5.3 Analysis and Description

With the data at hand do you think Ainu has a passive construction? Why/why not? Justify your answer in no less than 400 words making reference to the examples given in the sections above. If you would not call the construction under investigation 'passive', how else would you call it? Why?

**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

# <sup>6</sup> Relative Clauses

**Summary** 6.1 Background Information and Observation. – 6.2 Research. – 6.3 Analysis and Description.

#### 6.1 Background Information and Observation

In both HA and SA relative clauses are obtained via the gap strategy. The relativised noun in the matrix clause heads the relative clause (i.e it is the relative head, RH). There is no overt reference to the head noun within the relative clause (RC), so its original position is signalled only by a gap. There is no subordinator (e.g. a relative pronoun, a relativiser, …) and the right boundary of the relative clause is only marked by the position of the relative head.

Examples (1) and (2) show subject relativisation in HA and SA respectively.


The gap strategy is the only way to relativise verbal arguments (i.e. subject and objects) in both HA and SA. In line with the typological syntactic structure of OV languages (see Lesson 1), the relative clause always precedes the matrix clause.

*Consider this additional information…*

There exist other obliques in Ainu besides the ones expressing location discussed in Lesson 11. The following items are most commonly found: *ani* (HA and SA) that indicates instrument, *turano* (HA) or *tura* (SA) that indicates company, and *eun* (attested only in HA) that indicates the goal or beneficiary of an action. *Ani*, *tura(no)*, and *eun* follow the noun they refer to and are nothing but verb forms used adverbially – for instance, *ani* is a verb meaning 'carry' and *tura* is a verb meaning 'accompany'. The semantics of these verbs have led to their use to mark oblique case relations. Therefore, *ani*, *tura(no)*, and *eun* are different from postpositions like *-ta*, *-un*, *-(e)ne*, *-wa*, *-peka*, and *-poka* (see again Lesson 11).

#### *Dataset 1 – Relativisation of non-arguments*

Consider the following sentences showing relativised and non-relativised constructions. Why are the ungrammatical examples (marked with \*) so? What are the morphosyntactic differences with other instances of relativisation where ungrammaticality does not arise? Do you find any correspondence between these morphosyntactic differences and the function of the noun that is being relativised? On the basis of this, what amendments are you able to make to the information you started your analysis from?

#### **Set 1.1** (Hokkaidō Ainu)


9. *\* Katkemat hotke usi kar.*

'The woman prepared a [sleeping] place on which he [could] lie down.'

10. *Petotta cep poronno hemesu.*

'In the river many fish go upstream.'

11. *Acapo turano kuyupo kamuy cotca.*

'My elder brother shot the bear together with the uncle.'

12. *\* Cikir tanne kikir.*

'An insect whose legs are long.'

13. *Ani kuyupo kamuy tukan teppo.*

'The gun with which my older brother shot the bear.' (Bugaeva 2004, 95)

14. *Aekasihi ka nupe ranke.*

'My grandfather dropped tears (= was weeping).'

	- 'The men who came up to the doorway.' (Bugaeva 2004, 95)

'The insect's legs are long.'


#### **Set 1.2** (Sakhalin Ainu)


'The monster's upper jaw touched the sky.'


12. *Pon kahkemah* […] *'ene an 'ahciutah koitah.* '[Our] little young woman speaks to such old women.' (adapted from Dal Corso 2021)

#### 6.2 Research

Now look at this other dataset and also read the examples from other languages given after it, which you will need for the third and last activity of this lesson.

#### *Dataset 2 – Other cases*

Consider the following examples (? marks dubious cases). Given what you have discovered up to this point of your analysis, do you think these are relative clauses too? Why/why not? What is the morphosyntactic structure of these sentences? If you think these structures are not in fact relative clauses, how else would you call them?

#### **Set 2.1** (Hokkaidō Ainu)


#### **Set 2.2** (Sakhalin Ainu)


#### **Examples from Other Languages…**



Some languages may employ **retention** as a means to retrieve the original position of the relative head within the relative clause. Sometimes retention can be used **along with the gap strategy**, like in the case of Yoruba. In Yoruba retention is obligatory for both arguments and obliques. With arguments, we see **pronoun retention** which is when a pronoun is found within the relative clause to reference to the relativised noun. In the first sentence the resumptive pronoun is *ó*, that refers to the animate *ọkùnrin* 'man', while in the second sentence the pronoun is *náà*, that refers to the inanimate *aṣọ* 'cloth'. With obliques, on the other hand, the gap strategy is used, but we also see **retention of the preposition** that indicates the original function that the relative head had prior to relativisation. In the third sentence the relative head *àdá* 'cutlass' is gapped, but the preposition *fi* 'with' is retained within the relative clause.

#### Japanese (Japonic, Japan)



'The chopsticks with which I ate the fish.'

In some languages the gap strategy is the only available strategy for relativisation and it is used **regardless of the function** that the relativised noun had in the relative clause. Japanese, for instance, employs the gap strategy for both arguments and obliques. In the first sentence above, the relativised *hito* 'person' is an argument of the verb *atta* 'met' in the relative clause. In the second sentence, the relativised *hashi* 'chopsticks' is an oblique (instrument) of the verb *tabeta* 'ate'. Nevertheless, there is **no morphosyntactic indication** within the relative clause to signal this original function, which is then understood solely on a pragmatic basis.

#### Karachay-Balkar (Turkic, Russia) (examples from Comrie 1998, 81)

*[Prezident kel-gän] hapar.* president come-PTCP news 'The news that the president has come.'

*[Et biš-gän] iyis.* meat cook-PTCP smell 'The smell of meat cooking.'

Many languages also display constructions that, when we look at the linear order of constituents, are identical to relative clauses. In the Karachay-Balkar examples above the nouns *hapar* 'news' and *iyis* 'smell' follow a clause that they are the head of. This is exactly the order of constituents found in relative clauses in this language. However, both these nouns originally **do not occupy any positionwithin the relative clause** – both *kel* 'come' and *biš* 'cook' are intransitive verbs whose subject is already present. This means that nothing has been gapped from within the clause in square brackets. These constructions are not relative clauses, but rather they are treated as **general noun-modifying clause constructions** (or GNMCCs) (Comrie 1998). In GNMCCs a noun heads a clause while not being formerly included in it. There is **no explicit expression of the relation** between this head noun and the clause, but the clause simply describes the semantics of the head noun. We can think of the first sentence as 'What kind of news?' 'The president-has-come news.' Some languages employ a subordinator of some kind (like the participle *-gän* in Karachay-Balkar) but many others do not employ any overt marking of dependency.

#### 6.3 Analysis and Description

In no less than 400 words discuss relative clauses in HA and SA on the basis of your analysis of the data provided in Set 1 and Set 2.


**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

# <sup>7</sup> Negatives

**Summary** 7.1 Background Information and Observation. – 7.2 Research. – 7.3 Analysis and Description.

#### 7.1 Background Information and Observation

Negation in both HA and SA is expressed periphrastically. In HA the negative *somo* is placed before the verb (phrase) to negate it (Tamura 2000, 226).


In SA *hannehka* is found instead of *somo* (2) (Murasaki 1979, 109, 143). Moreover, *hannehka* can also be found to follow the verb (phrase) it negates (3) – in these cases it is preceded by *ka* 'even'.



*Consider this additional information…*

In line with the syntactic typology of the Ainu language (see Lesson 1) we expect adverbs to precede the verb they modify. This expectation is met, since adverbs are found in a pre-verbal position in both HA and SA.

#### *Dataset 1 – Possible negative structures*

Look at the following example clauses paying attention to ungrammatical ones (marked with \*). What are the possible syntactic layouts of negative constructions in HA and SA? What word classes are involved in these constructions and what word class do *somo* and *hannehka* belong to in your opinion? Motivate your answer. What examples do you find hard to analyse? What are the morphosyntactic characteristics that make these examples problematic?

#### **Set 1.1** (Hokkaidō Ainu)


'That noble man who had the village (= the village chief ) did it.' (Tamura 1985, 64)

	- 'You show me quickly.' (Tamura 1985, 18)
	- 'It doesn't have legs, hands nor bones…' (Tamura 1984, 56)


#### 7.2 Research

Now look at this other dataset and also read the examples from other languages given after it, which you will need for the third and last activity of this lesson.

#### *Dataset 2 – Constituent structure*

Consider the following negative clauses and pay attention to the nonnegative examples too. Only SA examples are provided for this dataset. What syntactic structures/processes are present in the non-negative clauses and how much of them can you find in the negative constructions? Review (if necessary) what word classes the constituents in the negative constructions belong to and provide a justification for your statements.

#### **Set 2** (Sakhalin Ainu)



#### **Examples from Other Languages…**

'Also there there was a big village.' (Dal Corso 2021)

Persian (Iranian, Iran) (examples from Neiloufar 2014, 26)


#### English (Germanic, UK)


Many languages possess periphrastic verb constructions made up of a verb and a non-verbal element (usually a noun, but also adjectives, adverbs or prepositions are common). Verbs such as 'do', 'take', and 'give' are most commonly used, but the choice of the verbs allowed to appear in these constructions is strictly language dependent and always limited to a certain number of elements. The verbal element in these constructions does not contribute any meaning to the whole phrase (or its meaning is downgraded), but it is the non-verbal element what contributes the core meaning to the expression. See for instance the Persian *guʃ kærdæn* where the noun 'ear' contributes the meaning of 'listening', or the English *do cleaning*  where the semantic content of the verb phrase is in the verbal noun *cleaning*. For this reason, the verb in these constructions is said to be a '**light verb**', in the sense that its original meaning is not retrievable as it would be when it is used alone and it only has the syntactic function of heading a verb phrase. The whole construction is referred to as '**light verb construction**' (see e.g. Butt 2010).

#### Malay (Austronesian, Malaysia)

(examples from Yap, Grunow-Hårsta, Wrona 2011, 13)


#### Korean (Koreanic, Korea)

(examples from Yoon 1991 in Yap, Grunow-Hårsta, Wrona 2011, 6)

*Chelswu-ka chayk-ul ppalli ilk-ess-um-i pwunmyengha-ta.* Chelswu-NOM book-ACC quickly read-PST-NMLZ-NOM evident-DECL 'It is evident (the fact) that Chelswu read the book quickly.'

**Nominalisation** is a process through which a verb, verb phrase or an entire clause, headed by a verbal element, becomes to function as a noun. Nominalisation can be a morphosyntactic or a semantic process. In the first case, the change of word category (from verb to noun) is overtly signalled by a morphosyntactic element, usually a **nominaliser**, like the Korean *-um*. The new nominal status of the phrase/ clause marked with *-um* is clear from the fact that the whole constituent can bear morphological elements that are only compatible with nouns (see the NOM suffix *-i*). Alternatively, nominalisation is not signalled overtly in any way morphosyntactically. In this case a verb phrase or an entire clause starts to function as a noun, like in the Malay case above where the verb phrase *makan lewat* 'eat late' functions as the argument of the main verb *bagus* 'be good'. Nominalisation in this case is also termed '**zero-nominalisation**', '**zero-derivation**' (e.g. Lieber 2005) or '**conversion**' (e.g. Bauer, Valera 2005).

#### 7.3 Analysis and Description

In no less than 400 words, list all constructions used in HA and SA to express negation providing an example for each construction.


**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

### <sup>8</sup> Noun Incorporation

**Summary** 8.1 Background Information and Observation. – 8.2 Research. – 8.3 Analysis and Description.

#### 8.1 Background Information and Observation

Noun incorporation (NI) is a process by which a noun stem, a noun root, or a noun phrase is combined with a verb to form one single stem. There have been many approaches to noun incorporation, some of which discuss it as a morphological process (e.g. Modena, Muro 2009) and some others that discuss it as a syntactic process (e.g. Baker 1988) or as semantic process (Chung, Ladusaw 2004). Traditionally, NI is said to be a valency-decreasing strategy, meaning that when the noun is incorporated in the verb this latter subcategorises for one less argument – if it is a transitive verb it becomes intransitive, if it is an intransitive verb already it becomes a zero-valency verb. After incorporation happens, the incorporated noun does not count as a verbal argument anymore, but it is part of the verb.

HA and SA showcase several cases of noun incorporation. In both varieties incorporation usually concerns the subject of intransitive verbs and the object of transitive verbs (i.e. the S and O arguments), while incorporation of the subject of transitive verbs is rare (Dal Corso 2021). Examples (1) and (2) illustrate S- and O-incorporation in HA, examples (3) and (4) illustrate the same in SA.

(1) *Sir-pirka.*

condition-be.good

'The condition is good.' = 'The weather is good.' (Tamura [1973] 2001, 119)

**Ca' Foscari Japanese Studies 18 Linguistics and Language Education 3 DOI 10.30687/978-88-6969-585-8/008**

(2) *E-wakka-ta kusu e-arpa.* 2SS-water-collect CAU.FIN 2SS-go.PL 'You go to draw water.' (UT 1987, line 457)

(3) *Sir-u-kunne.* condition-0-be.dark 'The condition is dark.' = 'It is night.' (Dal Corso 2021)


*Consider this additional information…*

Noun incorporation in Ainu is found to influence the stress pattern, which is then taken as a piece of evidence that incorporation has (not) occurred (Bugaeva 2004, 29). For example, in (1) stress falls only on *sir*, but in the non-incorporated synonymous version *sir pirka* both *sir*  and *pirka* would bear the stress. The stress pattern of *sirpirka* in (1) is therefore a signal that noun incorporation has happened.

*Dataset 1 – Evidence of incorporation*

Consider the following examples. Some feature noun incorporation and some do not (\* marks ungrammaticality). What morphosyntactic evidence do we have to say that noun incorporation has or has not occurred? What are the syntactic processes or the morphology that are insightful in this regard? Why does ungrammaticality arise?

**Set 1.1** (Hokkaidō Ainu)


'It is hot.' (Tamura 2000, 194)

6. *Ata wakka.*

'The water that You drew.'

7. *Orota eahun wa eapeare.* 'You enter there and light a fire.' (OS 1977, line 409)

8. *\* Taan wakka.*

'The water that You drew.'


#### **Set 1.2** (Sakhalin Ainu)

	- 'The house I got [for] myself.'

'The woman I married.'


#### 8.2 Research

Now look at this other dataset and also read the examples from other languages given after it, which you will need for the third and last activity of this lesson.

#### *Dataset 2 – Limits and peculiarities of NI*

Consider the sentences given below (\* marks ungrammaticality, ? marks dubious cases). Only SA examples are given for this dataset. Is there any syntactic or non-syntactic behavior that does not comply with what you have just observed about noun incorporation in Set 1? Why do you think ungrammaticality arise?

#### **Set 2** (Sakhalin Ainu)


#### **Examples from Other Languages…**

Mohawk (Iroquoian, USA and Canada) (examples from Baker 1996 in Muro 2009, 118)

*Thí kʌ ʌ-ye-nakt-a-núhweʔ-neʔ.* **that** FUT-3F.SG.S/3N.O-**bed**-LNK-like-PUNC 'She will like this bed.'

In some languages incorporation only allows noun roots to be incorporated into a verb. Some others also allow **larger nominals** to be incorporated (like noun stems made up of a noun root plus, for example, an adjective root) but do not allow these nominals to have modifiers (like quantifiers, demonstratives, relative clauses, …). This is not the case of Mohawk. In Mohawk a noun with **modifiers** can be incorporated, like we see for the noun *nakt* 'bed' in the example above. The modifier of this noun, here the demonstrative *thí kʌ* 'that', remains **stranded** (i.e. outside of the verbal form) but it still references the incorporated noun (Muro 2009, 108-25). The languages that allow this kind of incorporation usually impose **limitations on the syntactic kind** of the element that can remain stranded, so that not all syntactic constituents are acceptable in that position.

Hindi (Indo-Aryan, India)

(examples from Dayal 2011 in Borik, Gehrke 2015, 20)


In some languages noun incorporation is less strict syntactically. The incorporated noun is usually **less marked morphologically** (e.g. it cannot have markers of person, possession, number, …) but it **retains some syntactic freedom** – e.g. it can be separated from the incorporating verb by an adverb, a particle and the like. This never happens in noun incorporation proper, which is why these cases are defined as **pseudo-noun incorporation** (PNI) (Borik, Gehrke 2015). Pseudo-incorporated nouns too can have modifiers, like the adjective *puraanii* 'old' that modifies the pseudo-incorporated noun *kitaab* 'book' in the example above. However, PNI imposes some **restrictions in the semantics** of modifiers so that only modifiers with semantics that help deriving a verbal form describing an action with **cultural relevance** are acceptable. In Hindi, selling old books is recognised as an activity that belongs to the culture of the people speaking the language while, for instance, 'selling red books' is not. Therefore it would be impossible to substitute 'old' with 'red' in the example above. A different structure (one that does not involve PNI) must be used.

Chukchi (Chukotko-Kamchatkan, Russia) (examples from Polinskaja, Nedjalkov 1987 in Muro 2009, 3)

*ətləg-ən kawkaw-ək mətqə-rkele-gʔe.* father-ABS bread-LOC **butter-**spread.on-3SG.S 'Father spread butter on the bread.'

Blackfoot (Algonquian, Canada) (examples from Frantz 1971 in Mithun 1985, 858)

*Iihpokón-sskaawa nóko'sa.* **ball-**acquire.3SG my.child 'My child got a ball.'

*Nít-ohpokón-sskoawa nóko'sa.* I-**ball-**acquire.3SG my.child 'I provided my child with a ball.'

Noun incorporation may result in the **syntactic saturation** of the verb. For instance, if a noun is incorporated in a transitive verb, this verb becomes syntactically intransitive – i.e. it may take now just one argument, the subject. This is shown by the Chukchi example where the transitive *rkele* 'spread on' incorporates the noun *mətqə-* 'butter' and becomes intransitive. The new valency of the verb is clearly signalled by the **agreement** suffix *-gʔe*,that is used for marking third person subject on intransitives.

In other languages noun incorporation does not necessarily result in the syntactic saturation of the verb. In the first example from Blackfoot we see object incorporation of the noun *iihpokón-* 'ball' into the verb *sskaawa* 'acquire' which causes syntactic saturation – the only argument the verb can take is now the subject, here *nóko'sa* 'my child'. However, this same incorporation **may not cause syntactic saturation** and the argument slot left vacant by the incorporated noun is filled by a new nominal or by an already present one. In the second example from Blackfoot *nóko'sa* 'my child' takes the empty slot of the object, it leaves the subject slot empty and so the verb can now take a new subject, here *nít* 'I'. The verb **has not changed in valency** as it is still transitive. With this kind of incorporation (with no syntactic saturation) the semantics of the verb **may slightly change**. Furthermore, many languages impose **restrictions on the semantics** of the noun that can replace the incorporated noun.

#### 8.3 Analysis and Description

In no less than 400 words discuss noun incorporation in HA and SA.


**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

# <sup>9</sup> Source of Information

**Summary** 9.1 Background Information and Observation. – 9.2 Research. – 9.3 Analysis and Description.

#### 9.1 Background Information and Observation

In both HA and SA there exist a number of sentence-final expressions that indicate how the speaker has obtained the information that is the content of her utterance and what was the source of this information. In other words, these final forms indicate the kind of evidence the speaker has in order to say what she says. The HA *siri ne* in (1) and the SA *humihi 'an* in (2) provide an example. Additional background context information is given in square brackets.

(1) *Sinuma ka ko-ray-niwkes siri ne (noyne iki a).* he even APPL-3SS/3SO/die-be.difficult **siri ne** as.if AP.do PRF 'It was (like) he couldn't [separate from me] either.' (Kayano 4-4,10) [The speaker is being hugged by her husband and sees that he cannot let go of her since he is too happy]

In (1) the speaker sees first-person that her husband is having trouble separating from her because he misses her.


In (2) the speaker realises that the people next to him are completely still and, on the basis of that, deduces that they must be sleeping deeply.

Since these are sentence-final forms they come right after a verb. Their use is not obligatory on every sentence.

	- One of the ways to express possession in HA and SA is with the suffix *-(V)hV* that is attached to the possessed noun (see Lesson 4).
	- The alloform *-hV* is attached to noun roots ending in a vowel and the alloform *-VhV* is attached to noun roots ending in a consonant.
	- The vowel(s) of this suffix is copied from the last vowel present in the noun root to which the suffix attaches. Therefore we have for instance *sapa* 'head' > *sapaha* or *tek* 'hand' > *tekehe*.
	- Sometimes the final segment *-hV* of the possessive suffix is truncated (e.g. *sapaha > sapa* or *tekehe > teke*).
	- For some consonant-final noun roots this vowel-copying rule does not apply and the vowel in the suffix is different from the last one present in the root, resulting in an 'irregular' form – e.g. *par* 'mouth' > *paroho*.
	- On this point HA and SA may differ as some 'regular' possessive forms in HA are 'irregular' in SA, like for example *tek* 'hand' > *tekehe* (HA) but *tek* 'hand' > *tekihi* (SA).
	- The possessor on these forms is marked again on the noun via a transitive personal agreement prefix in the subject form – e.g. *kusapaha* 'my head' (HA, SA), *eparoho* 'your mouth' (HA), or *tekihi* 'his/her/its/their hand' (SA).
	- When the possessor is a third person it can be expressed by a separate noun that always precedes the possessed noun – e.g. *seta paroho* 'the dog's mouth'.

#### *Dataset 1 – Variety of forms*

Among the following example clauses there are some that include other sentence-final expressions like the ones we saw in (1) and (2) above and some others that do not (additional examples are given for you to infer the meaning of words). Besides *siri ne* and *humihi 'an* what are the other forms present in HA and SA? Make a list of them. Then make reference to the other examples to understand what is the morphological structure of the forms you have listed and what word classes are involved. After you are done with the morphological analysis return to the examples that include the sentence-final forms – can you make any observations on their syntactic structure? Background context information is given for these examples, but you will need it later for Set 2.

#### **Set 1.1** (Hokkaidō Ainu)

1. *Sinuma ka korayniwkes siri ne (noyne iki a).*

'It was (like) he couldn't [separate from me] either.' (Kayano 4-4,10)

[The speaker is being hugged by her husband and sees that he cannot let go of her since he is too happy]

2. *Areska wa an.*

'I had raised her [so].' (Bugaeva 2004, 307)


'It seemed my brother was coming towards me.' (adapted from Tamura 1985, 60)

[A stoat, character of the tale, turns himself into a woman and from afar sees his brother walking in his direction]


'It seems that my degenerate daughter has had too much of a violent temperament.' (Tamura 1984, 34)

[The speaker comments like this after the main character of the story recounts the difficult travel he had to make because his mother, the speaker's daughter, had told him to].


'It seemed I was carried and going somewhere.' (Tamura 1985, 4) [A woman is inside a coffin and from the movements and the voices coming from outside she figures out she is being carried some place up a hill]

	- '[My father] jumped and stood up.' (Bugaeva 2004, 204)

'A young man, who I don't know, has come.' (Bugaeva 2004, 256)

[A young woman says this to the owner of the house after having met an unknown man at the door]

13. *Apoho ene hawe ne ciki, isamun ek.*

'If you are my child, come next to me.' (Tamura 1984, 40)

[A woman says this after a young boy introduced himself to her saying that he was her son]

14. *Tanto sirpirka.* 'Today the appearance is good (= the weather is nice).'


'[The monster] ended up eating […] even all the people.' (Dal Corso 2021)


#### 9.2 Research

Now look at this other dataset and also read the examples from other languages given after it, which you will need for the third and last activity of this lesson.

*Dataset 2 – Structure and meaning*

Consider the sentences given here together with the ones in Set 1 for which background context is given. Do you find any variants of the sentence-final forms you have just listed? Why are the sentences marked with \* ungrammatical? What are the alternative forms you see and what changes in their morphosyntactic structure? On the basis of background context can you explain what is the meaning(s) and function(s) of the different sentence-final forms? How can they be subdivided within each variety and what are the differences between HA and SA?

#### **Set 2.1** (Hokkaidō Ainu)

1. *Aonaha heru earanukar kor ora ihoppa siri an?!*

'Only once I see my father and then he leaves me?!' (adapted from Kayano 1-8,12) [The speaker utters this sentence after the father he hadn't seen for a long time walks away from him]

2. *Ukuran ka yaanipo isam* […] *sirki.*

'Even in the evening it seemed he was almost […] dying.' (Tamura 1984, 14) [The speaker tells about the time she healed a sick person and speaks about their physical conditions]

3. *Okkayo haw* […] *"ahunke yak pirka wa" sekor hawas.* '"You may let him come in" said the voice [of] a young man.' (Bugaeva 2004, 257) [The speaker hears the woman who just met him at the door say this once she goes back in to inform the owner of the house about who just arrived]


'It seemed they were sleeping while snoring loudly.' (Dal Corso 2021) [The speaker is lying nearby some people who are sleeping] 4. *'Usahpa 'usahpa cisehehcin'onne (…) 'ampahci.*

'[The man] took [the bears] one by one to their house (=cage).' (Dal Corso 2021)


#### **Examples from Other Languages…**

Ngiyambaa (Pama-Nyungan, Australia) (data from Donaldson 1980 in Aikhenvald 2004, 34)

*Ŋindu-gara girambiyi.* you.NOM-SENS.EV sick.PAST 'You were sick (one could see it).'

*Dhagun-gir-gara ŋina dhiŋgaː ga-ɽa.* earth-NASTY.WITH-SENS.EV this.ABS meat.ABS be-PRES 'This meat tastes nasty with earth (I have tasted it).'

Foe (Austronesian, Papua-New-Guinea) (data from Rule 1977 in Aikhenvald 2004, 62)

Aiya bare wa-boba'ae. plane come-EV.VIS 'A plane is coming (I can see it).'

Many languages of the world have a system of forms dedicated to indicate how the speaker acquired the information that is the content of what she says and, additionally, what was the physical or non-physical means through which this acquisition happened. These forms are called **evidentials** and the category to which they belong is called **evidentiality** (e.g. Aikhenvald 2004). One type of evidentials are the ones that indicate that the content of information is acquired directly (or first-hand) by the speaker – these are usually called **direct evidentials**. A language may not have separate direct evidentials depending on the kind of source through which information was acquired, like Ngiyambaa where the form *-gara* can be used regardless of the stimulus at the basis of the acquisition (in the first sentence the speaker knows that 'you were sick' from sight but in the second sentence she knows that 'the meat tastes nasty' from taste). Direct evidentials may otherwise differ according to the stimulus at the basis of the acquisition, like in Foe where the form *-boba'ae* specifically indicates that information is acquired directly through a visual stimulus.

#### Amdo Tibetan (Tibeto-Burma, China) (data from Sun 1993 in Aikhenvald 2004, 45)

*Tʂaɕʰi=kə ʰtæ ȵu=zəg.* Bkra-shis=ERG horse buy=INFR 'Bkra-shis bought a horse (speaker inferred it).' Another kind of evidentials are the ones that indicate that information was acquired indirectly – these are called **indirect evidentials** or also **inferentials**. In many languages of the world the stimulus or means of acquisition of inferentials is left unspecified, like in Amdo Tibetan. In the sentence above the inferential *=zəg* only indicates that acquisition of the fact that 'Bkra-shis bought a horse' happened indirectly, but does not specify whether this happened through a visual stimulus (i.e. the speaker seeing that Bkra-shis built a stable), an auditive stimulus (i.e. the speaker hears a horse neighing) or else. In this case only context may provide clarification.

Lezgian (Caucasic, Azerbaijan) (data from Haspelmath 1993 in Aikhenvald 2004, 31)


Finally, some languages have evidentials dedicated to mark verbal report – these are called **hearsay evidentials**. Hearsay evidentials mark that the information content of the utterance was acquired from someone else and it is being reported as it is. Some languages may distinguish **quotatives** and **reportatives** within hearsay evidentials. With quotatives the speaker reports an information repeating, or somehow 'quoting', the person who originally said those words. With reportatives the speaker reports an information indirectly, most times without specifying the original source of the statement. The Lezgian *-lda* is used above in this latter function.

#### 9.3 Analysis and Description

In no less than 400 words, list all forms used in HA and SA to mark source of information (if present, indicate the variants of one same form). Making reference to the examples in the sets above, describe in your own words the meaning(s) and function(s) of each form. If necessary, make reference to the examples from other languages in your argumentation.


**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

### <sup>10</sup> Antipassive

**Summary** 10.1 Background Information and Observation. – 10.2 Research. – 10.3 Analysis and Description.

#### 10.1 Background Information and Observation

Antipassivisation is a process common in, but not exclusive to, ergative languages and it usually concerns transitive and ditransitive verbs. In a prototypical antipassive construction an original object argument of the verb is demoted to be an oblique (Payne 2006, 255). At discourse level, antipassivisation serves to de-focus an element that has low referentiality in the context. Antipassivisation is therefore similar to passivisation, the difference between the two processes being that the former downplays the centrality of an object while the latter downplays the centrality of the subject.

In both HA and SA there exists the antipassive marker *i-*. This verbal prefix is formally identical to the fourth person object marker *i-* (see Lessons 3 and 4). However, the origin of these two prefixes is diachronically distinct, with the antipassive marker being historically older than the 4O agreement prefix. A piece of evidence to argue for the older origins of the antipassive marker comes from phonology since, for reasons connected to a process of lexicalisation of the antipassive form, antipassive *i-* triggers glide insertion with vowelinitial verb stems (see Lesson 3). The 4O agreement prefix is usually not found to cause glide insertion (Tamura 1970; Bugaeva 2018, 7).

Examples (2) and (4) show an antipassive construction in HA and SA, and examples (1) and (3) respectively show their non-antipassive version.

> **Ca' Foscari Japanese Studies 18 Linguistics and Language Education 3 DOI 10.30687/978-88-6969-585-8/010**


The proposed path of development of the Ainu antipassive marker is as follows (Bugaeva 2018).


#### *Dataset 1 – Morphosyntactic constraints*

Examples (1)-(4) above already show a behavior of the Ainu antipassive that is at odds with the characteristics of a prototypical antipassive construction. Consider the following examples (\* marks ungrammatical examples). What is the morphosyntactic structure of the Ainu antipassive construction? What causes the ungrammaticality of examples marked with \*? Does antipassivisation affect verbal valency? How? What is the morphological evidence for that?

#### **Set 1.1** (Hokkaidō Ainu)

	- 'We heard a news.'

#### **Set 1.2** (Sakhalin Ainu)


#### 10.2 Research

Now look at this other dataset and also read the examples from other languages given after it, which you will need for the third and last activity of this lesson.

*Dataset 2 – Antipassive's semantics*

Consider the following examples that illustrate the semantic characteristics of the Ainu antipassive and the relative translations. Judging from the grammatical and ungrammatical examples given here, how is the semantic reference of the antipassive restricted? Do the semantic restrictions of the applicative affect the meaning of the verb form as a whole? How? Is there any difference between HA and SA? Are there any instances where the antipassive causes a morphological change of the non-antipassive base verb?

#### **Set 2.1** (Hokkaidō Ainu)

	- 'As I raised my head and looked around …' (OS 1980, line 160)

'I ate and went to sleep.' (OS 1980, line 552)

	- Intended meaning: 'He made the birds go to their nests.'

'You saw me in the shape of a dog.' (OS 1979, line 669)

13. *Sake aku ka eramiskari.* 'I had never drunk sake.' (IH 1986, line 482)

#### **Set 2.2** (Sakhalin Ainu)



#### **Examples from Other Languages…**

Central Alaskan Yup'ik (Eskimo-Aleut, USA) (examples from Miyaoka 2012)


A **prototypical antipassive** construction has three formal characteristics (Payne 2006, 255): 1) the A argument (subject of transitive verb) **becomes the S argument** (subject of intransitive verb), which may be clear from a **morphological change** in the person agreement forms used on the verb, 2) the originally transitive **verb becomes intransitive**, which makes antipassivisation a **valency-changing** (valency-decreasing) strategy, 3) the original object of the transitive verb appears in an **oblique case**. The Central Alaskan Yup'ik examples above illustrate a prototypical antipassive. After antipassivisation is obtained via the suffix *-i*, the formerly transitive verb becomes intransitive – this is evident from person agreement as in the first example we find the transitive agreement *-aa* (referencing to a third person subject acting on a third person object), but in the second example we find the intransitive agreement *-uq* that only references a third person subject. The subject in the antipassive construction is an S – this is evident from the absolutive null marking on *angun* 'man' in contrast to the relative-ergative suffix *-m* in the non-antipassive sentence. Finally, the original direct object *kuvya* 'net' is expressed in the antipassive construction via an oblique (here with the ablative case in *-minek*).

Puma (Sino-Tibetan, Nepal) (examples from Bickel, Gaenszle 2015)

*Som-kha-mʌ-tuk.* love**-ANTIP**-3PS-love.NPTS 'They love people.'

Antipassives can have their origin in **noun incorporation**. The antipassive *-kha* of Puma originated from a no longer existent noun meaning 'all' which was incorporated in the verb and later developed the function of referencing to a **generic object**. In the example above, the antipassive on the verb *som* 'love' references to 'people' in general. The peculiarity of antipassives that originate from noun incorporation seems to be that the construction **does not allow** the demoted object to be expressed overtly via an oblique. In languages possessing this kind of applicatives the demoted object is **always omitted**.

Tlachichilco Tepehua (Totonac-Tepehua, Mexico) (examples from Watters 2017)

*Sa:-y Sa:-nan* hit-IPFV hit.IPFV**-ANTIP** 'S/he hits him/her/it.' 'S/he plays music.'

*Mispa:-y Mispa:-nan* know-IPFV know.IPFV**-ANTIP** 'S/he knows him/her/it.' 'S/he knows the location.'

Antipassivisation often results in **lexicalisation**. This happens more commonly in those languages (like Puma above) whose antipassive construction does not allow the overt expression of the demoted object. Tlachichilco Tepehua represents one more case of such languages. In fact, once the antipassive is obtained via the suffix *-nVn*, object expression is blocked and the antipassive takes a **restricted set of objects** or even only **one object specifically** as its semantic referent. This often causes a **change of meaning** in the antipassivised verb form compared to the one of the non-antipassive verb. Therefore, for instance the verb form *sa:nan* from *sa:* 'hit' only means 'play music', with the antipassive referencing specifically to musical instruments.

#### 10.3 Analysis and Description

Discuss, in no less than 300 words, the morphosyntactic and semantic characteristics of the Ainu antipassive. Cite relevant examples from Set 1 and Set 2 where needed.


**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

# <sup>11</sup> Motion and Location

**Summary** 11.1 Background Information and Observation. – 11.2 Research. – 11.3 Analysis and Description.

#### 11.1 Background Information and Observation

In both HA and SA motion and location are usually obliques (more rarely they appear as arguments). Morphosyntactically motion and location are expressed via a construction involving the noun *or* (HA), *oh/oro* (SA) 'place'. This noun follows another noun that indicates a specific place and it is in its turn followed by a locative postposition that indicates location or specifies the kind of motion involved. One example of such postpositions is *-wa*, expressing motion away from a place, that we have already encountered in the impersonal-passive construction (cf. Lesson 5). Examples (1) and (2) illustrate the construction with *-orwa/-orowa* in HA and SA respectively.


Besides *-wa* there exist other locative postpositions. In HA we find *-ta* that indicates stative location, *-un* that indicates motion towards

> **Ca' Foscari Japanese Studies 18 Linguistics and Language Education 3 DOI 10.30687/978-88-6969-585-8/011**

a place, and *-peka* that indicates motion through a place (Tamura 2000, 104-8). In SA we find *-ta* that indicates stative location, *-(e)ne* that indicates motion towards a place, *-peka* that indicates straight motion through a place, and *-poka* that indicates irregular motion through a place. Sometimes, depending on the semantics of the verb and on the kind of place where motion happens, *-peka* and *-poka* can be interchangeable (Murasaki 1979, 118-19; Dal Corso 2021).

*Consider this additional information…*

The HA postposition *-un* has its origin in the transitive verb *un* 'be in (a certain place)'. The original meaning of this verb is essentially stative, therefore its development into a postposition indicating motion is not clear.

(3) *Ya un kur.* mainland 3SS/3SO/**be.in** person 'A person who is in the mainland (= who lives/comes from the mainland, i.e. an Ainu).'

The same verb is also present in SA, but the use of *-un* as a locative postposition in this Ainu variety is much more limited. In fact it is used only with the meaning of 'from' in expressions that indicate the origins of someone or something.

(4) *Tokyo-'oro-'un 'aynu*. Tokyo**-place-in** person 'A person from Tokyo.' (Murasaki 1979, 99)

#### *Dataset 1 – Semantic complexity of motion and location*

Consider the following examples paying attention to ungrammatical ones (marked with \*). What are the phonological processes occurring on the noun *oro* 'place' when used with the various locative postpositions (cf. Lesson 3)? Do you notice any discrepancies in the use and/or meaning of locative postpositions with regards to the information you have been given? What makes the ungrammatical examples so?

**Set 1.1** (Hokkaidō Ainu)




#### 11.2 Research

Now look at this other dataset and also read the examples from other languages given after it, which you will need for the third and last activity of this lesson.

#### *Dataset 2 – Intrinsically locative nouns*

Consider the following examples. Only SA examples are given for this set. What new items expressing location do you see used as substitutes of *oh/oro*? Are these all nouns? If not, what word class do you think they belong to? Motivate your answer. What kind of locative relation do they express (under, over, next to, …)? Try to find the semantic difference among the ones with a similar meaning. How do locative constructions change morphologically when the referent of the locative expression is not a third person?

#### **Set 2** (Sakhalin Ainu)


3. *Nupuru tonkeene* […] *hemata puyehe 'ahun.* 'In the center of the mountain [side] […] some kind of hole opened.' (Dal Corso 2021) 4. *Ruu kaari makan.* 'He went uphill along the trail.' (Dal Corso 2021) 5. *\* Apa kuhekota.* 'I went towards the door.' 6. *Ota poyehci teh 'otatunketa 'ahunkehci.* 'They dug [holes in] the sand and put [praying sticks] in the center of the sand.' (Dal Corso 2021) 7. *Kuy naa yayuh naa renkayne tocaapeka 'okay.* 'There were a lot of larixes and Sakhalin firs too along the lake's shores.' (Dal Corso 2021) 8. *\* Ruu kaaripeka makan.* 'He went uphill along the trail.' 9. *'Orohko 'ahci sineh 'iyohta* […] *'an.* 'One old Orok woman lived […] at my place.' (Dal Corso 2021) 10. *Tooro hekota cas.* 'It ran to Tooro.' (Dal Corso 2021) 11. *Cipihisanketa neeroh kamuyutah yapahci.* 'Those gods surfaced [from the sea] next to his boat.' (Dal Corso 2021) 12. *Enhekota apkas.* 'He walked towards me.' 13. *Kenemoto'owsiketa hemata ka kamuy he 'an.* 'Under an alder there was some kind of animal.' (Dal Corso 2021) 14. *\* 'Itanki* […] *ansanketa ama.* 'He placed the bowl […] next to me.' 15. *Rerannoskekepoka 'anciw.* 'I [had] pierced him through the center of [his] chest.' (Dal Corso 2021) 16. *'Enko utah neampe* […] *'Usoro'onne 'oman.* 'Half of the people […] went towards 'Usoro.' (Dal Corso 2021) 17. *'Inoskeketa 'an kamuy reekoh 'aspekoro kamuyhenke.* 'The animal that was in the center [of the group was] an old god with mighty big back fins.' (Dal Corso 2021) 18. *'Eyooko utah rawta 'okay.* 'The people who lay in wait were under [the slope].' (Dal Corso 2021) 19. *'Itanki* […] *isanketa ama.* 'He placed the bowl […] next to me.' (Dal Corso 2021) 20. *Niitonke tuye teh horahte.*

'He cut the center of the tree and felled it.'

#### **Examples from Other Languages…**

#### Finnish (Uralic, Finland)


Finnish has a complex system of grammatical cases that distinguishes **locative cases** into **internal** and **external**. Internal locative cases indicate state inside (inessive), motion out of (elative), or motion into something (illative). External locative cases indicate state on (adessive), motion away from (ablative), or motion onto something (allative). The use of these cases can be extended to express meanings that are not locative *per se*. For instance, the inessive is used to express punctual time (e.g. 'in January') and the adessive is used to express, among other things, means and instrument (e.g. 'by train').

#### Icelandic (Germanic, Iceland)


In Icelandic the prepositions *í* and *á* are used to express state and motion to some place. The choice of which one of these prepositions to use is determined on the semantics of the noun they accompany. The reading of the locative expression as a state or as a motion, on the contrary, is clear from **grammatical case** (besides being possibly clear from the verb's semantics). In fact, the accusative case is used with *í* and *á* to express motion and the dative case is used to express state.

#### 11.3 Analysis and Description

Describe, in no less than 400 words, the ways to express motion and location in SA drawing from the examples given in Set 1.2 and Set 2. Be specific about the morphosyntactic structure of locative constructions and about the word classes involved. Provide a motivation for your statements.


**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

### <sup>12</sup> Aspect

**Summary** 12.1 Background Information and Observation. – 12.2 Research. – 12.3 Analysis and Description.

#### 12.1 Background Information and Observation

Ainu counts numerous markers for aspect and mood (e.g. Murasaki 1979; Tamura 2000; Bugaeva 2004; 2012). The language, however, has no dedicated markers to express tense, that is therefore unmarked as a verbal category. Generally, aspectual and modal constructions are formed periphrastically with the aspectual or modal marker following a notional verb, as it can be seen in the examples below. Aspectual or modal markes show different stages of grammaticalisation, so the morphological complexity featured by predicates that are marked for aspect or mood varies from case to case (e.g. Dal Corso 2020).

Example (1) shows *kusu ne*, expressing intentional mood in HA. This modal expression is made of the causal-final linker *kusu* and the copula *ne*. Example (2) shows *koyaykus* 'not be able', expressing incapability in SA. The modal verb *koyaykus* forms with the preceding notional verb a light verb construction (Dal Corso 2020) (see Lesson 7).

(1) *Yakun ku-ye wa eci-nu-re kusu ne.* if 1SS-3SO/say and 1SS>2SOI-3SO/hear-CAUS **INTN** 'If [so], I'm going to tell it and make you hear (= I'm going to tell you).' (Tamura 1984, 18)

> **Ca' Foscari Japanese Studies 18 Linguistics and Language Education 3 DOI 10.30687/978-88-6969-585-8/012**

(2) *Pirika nispa-utara ne yahka ray-ki koyaykus.* be.good noble.man-COLL COP though 3PS/3SO/die-TR **SLV/VO/not.be.able** 'Even the noble rich men could not kill it.' (Pilsudski 1912, 240)

The above examples serve only as an illustration. In this lesson you will be focusing on aspect only.

#### *Consider this additional information…*

Although there exist no formal devices that have the primary function of expressing tense, Ainu is certainly not a tenseless language. In most cases, a predicate's relative tense reference (present, past, or future with respect to speech time) is clear from context or it may be indicated by the use of time adverbs. Furthermore, aspect has been often recognised as having semantico-pragmatic extensions to encode tense (e.g. Kindaichi 1931; Tamura 2000, 111; Dal Corso forthcoming). This means that the use of certain aspectual forms helps put separate events into a logical relation with each other and, by consequence, they highlight the temporal relation of said events. The same semantico-pragmatic extensions have been observed for evidentiality (Dal Corso 2018) (see Lesson 9). In Ainu, time reference (i.e. relative tense) is therefore essentially a by-product of the use of other verbal categories.

#### *Dataset 1 – Types of aspect*

Consider the following examples featuring three different aspectual forms of HA and SA. Examples not including aspectuals are also included for you to extrapolate the meaning of words. What are the three aspectual forms found in the dataset? What is the meaning of the words that constitute them? What type of aspect to they encode? Do you notice any morphological change among instances of the same aspectual form? Of what kind? How can you account for them? How do aspectual markers with the same function differ formally between HA and SA?

#### **Set 1.1** (Hokkaidō Ainu)


'The owner of the house was walking around outside.' (Bugaeva 2004, 319)

	- 1. *Oha kemihi pateh* […] *pon nay neeno an teh netopakihi isam.* 'There was only his blood […] [flowing] like a little river and his corpse was not [there].' (Pilsudski 1912, 110)
	- 2. *'Ehankeno 'an pe ka 'emuyke kehke wa cokoko wa 'isam.* '[In the confusion of the battle, the monster] ended up breaking and felling even all the things (= the trees) around.' (Dal Corso 2021)
	- 3. *Soyta oka utah temana kihci kusu okayahci?* 'How were the people outside doing?' (Dal Corso 2021)
	- 4. *'Onnewrah ramrenkayne uyna wa eyhoh.* 'He picks up eagle's feathers as he wishes and sells them.' (Dal Corso 2021)
	- 5. *Hosipihci nukarahci kusu neeroh hekacita kanna sapahci.* 'Since they saw them returning, those boys went back down [to the village.]' (Dal Corso 2021)
	- 6. *'Otakata 'ampene ruwehehcin maake teh an.*
	- '[The animals] had clearly left their footprints on the beach.' (Dal Corso 2021) 7. *Kemaha ka 'okore tuyteh wa 'isam.*
		- 'Even her legs ended up falling off.' (Dal Corso 2021)
	- 8. *Susucayteh'orowa hemata rusasoma rahki kusu an.* 'Some kind of straw bundle was hanging from the willow's limbs.' (Dal Corso 2021)
	- 9. *'Ahciutah na henkeutah na 'okayahci.* 'There were both old women and old men.' (Dal Corso 2021)
	- 10. *Tumi kihci kusu suy taata sapahci.* 'They went down there again to make war.' (Dal Corso 2021)
	- 11. *Nah kii yayne 'ampene pakakara teh 'an.* 'She acted [like] so and eventually she got completely crazy.' (adapted from Dal Corso 2021)

#### 12.2 Research

Now look at this other dataset and also read the examples from other languages given after it, which you will need for the third and last activity of this lesson.

#### *Dataset 2 – Functional overlap and semantic differences*

Consider the following examples that feature two more aspectual markers of HA and SA respectively. Are there other markers in the two varietes that are similar in function to the new ones introduced here? How would you explain this variety of forms to express the same (or a similar) function? Are there any uses of aspectual markers that remain difficult to account for given the data provided?

#### **Set 2.1** (Hokkaidō Ainu)


	- 'Then we finished greeting each other and …' (Bugaeva 2004, 259)

15. *Tumunci kamuy umurek wa oka.* 'The evil gods were married.' (Tamura 1984, 20)

#### **Set 2.2** (Sakhalin Ainu)


#### **Examples from Other Languages…**

English (Germanic, UK)

(examples from van Lambalgen, Hamm 2005, 90-6)

*Know Push Push a cart Reach the top Build a house Flash*

Verbs in any given language differ strikingly one from the other not only with regards to their pronunciation, syntactic valency, etc., but also with regards to the **type of event(s)** they denote – i.e. their **core semantics**. Events in the real world do not have all the same properties: some can be prolonged, some cannot, some subsume a change of some kind, some others do not. The English verbs above serve to illustrate the verb classes usually distinguished in languages according to the properties of the denoted event. The verb 'know' is here an example of a **state verb** – it denotes a type of event which subsumes no change nor is initiated by a force of some kind. The verb 'push' alone is considered an **action (in the narrow sense) verb** – it denotes a type of event initiated by a force (e.g. someone starting a motion) but is subsumes no change whatsoever. In the predicate 'push a cart', on the contrary, the verb 'push' is an **action (in the wide sense) verb** – it denotes a type of event that is initiated by someone and that subsumes some kind of change (in this case, a change of position in the place where the cart is). The verb 'reach' in 'reach the top' is called an **achievement verb** – it denotes a type of event subsuming a change that happens suddenly and that usually implies 1) that the situation present before (here someone's being on their way to the top), which culminates with the sudden change, ceases to exist and 2) the achievement of a new state of things (here being at the top of some place). 'Build' in 'build a house' is referred to as an **accomplishment verb** – it denotes a type of event initiated by someone, that subsumes some kind of change (here the house being built), that culminates (gradually) with a change (the finished house), and ends with the achievement of a new state of things (a house where before there was nothing). Finally, 'flash' is referred to as a **semelfactive verb** – it denotes a type of event that only has a culmination, but lacks a force that initiates it, a resultant state, and any kind of change. These are the six classes of *Aktionsart* outlined by van Lambalgen, Hamm (2005), which include the four classes originally distinguished by Vendler (1967). Defining the *Aktionsart* of a predicate is useful to understand its **incompatibilities with certain categories or lexical items** (e.g. time adverbs, aspect, number, definiteness, …) that, with their function or semantics, affect the internal structure of the event. It is important to remember that the *Aktionsart* of verbs that are synonymous in different languages may be very different – that is, not all languages **perceive events the same way**. Moreover, one verb of a language may belong to **more than one** *Aktionsart* **class**, depending on the type and semantics of its arguments and adjuncts.

#### Japanese (Japonic, Japan)


Intended meaning: \* 'A person has drunk coffee.'

One category that is found to **interact closely with** *Aktionsart* is aspect. In various languages certain types of aspect cannot be used with predicates/verbs belonging to certain *Aktionsarten* because the properties of aspect are **incompatible with the event structure** denoted by the predicate/verb. In Japanese the verb *kowareru*  'break' is a static (instantaneous) verb, which means that the event of breaking is perceived as not possibly having a duration – a teacup is either broken or it is not. On the contrary, *nomu* 'drink' is a dynamic verb as it denotes an event with a duration – the event persists all the time a person takes to drink and finish coffee. The auxiliary form *-te iru* is employed in Japanese to express both progressive and resultative aspect. **The reading as one or the other** is decided on the basis of the *Aktionsart* of the predicate. Japanese shows a case where an aspectual form can be used with different *Aktionsarten* and takes different readings, but in many languages it can be the case that a certain aspectual form is simply **completely unacceptable**.

#### 12.3 Analysis and Description

Describe, in no less than 300 words, the HA aspectual markers *wa an/ okay* and *wa isam*, and the SA aspectual markers *teh an* and *wa isam*. Draw from the examples in Set 1 and Set 2 where needed.


**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

# <sup>13</sup> Applicatives

**Summary** 13.1 Background Information and Observation. – 13.2 Research. – 13.3 Analysis and Description.

#### 13.1 Background Information and Observation

Applicatives are a valency-changing strategy present in many languages. In an applicative construction what would be normally expressed with an oblique is promoted to be an argument of the verb (Payne 2006, 264-5). As such applicatives usually function as a referent-focusing strategy at discourse level, because they bring into focus an otherwise peripherical element. An applicative construction is most commonly formally signalled via dedicated morphosyntax on the verb.

In HA and SA we find three applicative prefixes: *e-*, *ko-*, and *o-*  (Tamura 2000, 206-8; Murasaki 1979, 32-4; Dal Corso 2021; Bugaeva 2010). These applicative morphemes are attached to the verb and have linear priority over personal agreement prefixes. Examples (1) and (2) show an applicative construction in HA and SA, while examples (3) and (4) show the relative non-applicative constructions.

(1) *A-kor yupo a-ko-itak.* 4S-3SO/have **older.brother** 4S-**APPL-3SO/**speak 'I said [so] to my older brother.' (OS 1980, line 11)

(2) *Suma kema 'e-tuhse-ka.* stone **leg APPL-**3SS/**3SO/**3SOI/jump-TR 'He/she kicked the stone with [his/her] foot.' (Murasaki 1979, 32) (3) *A-kor yupo eun itak-an.* 4S-3SO/have **older.brother towards** speak-4S 'I said [so] to my older brother.' (4) *Suma kema ani tuhse-ka.* stone **leg with** 3SS/3SO/jump-TR 'He/she kicked the stone with [his/her] foot.'

*Dataset 1 – Applicatives and thematic roles*

Consider the following examples featuring the applicatives *e-*, *ko-*, and *o-* (\* marks ungrammatical examples). The non-applicative variants of some of these examples are also included. What causes the ungrammaticality? Is there any correspondence between the properties of the noun promoted to argument of the verb and the applicative form used? Can you point out what determines the choice of the applicative form?

**Set 1.1** (Hokkaidō Ainu)

	- 'He shot with a cannon.' (Bugaeva 2010, 768)

#### **Set 1.2** (Sakhalin Ainu)


'Since the houses [of the village] of his younger brother were on the opposite bank [of the river] …' (Dal Corso 2021)

9. *\* Suma kema kotuhseka.* 'He/she kicked the stone with [his/her] foot.'

10. *\* Nokanramuhu cisehehcin neya 'arapesca koan kusu …* 'Since the houses [of the village] of his younger brother were on the opposite bank [of the river] …'


#### 13.2 Research

Now look at this other dataset and also read the examples from other languages given after it, which you will need for the third and last activity of this lesson.

*Dataset 2 – Polysemy and polyfunctionality*

Consider the following additional examples. Given what you have discovered up to this point of the analysis, what is peculiar about the use of applicatives in these instances? What new functional extensions and/or semantic properties of applicatives do you notice? In light of this, how can you refine your account of HA and SA applicatives?

#### **Set 2.1** (Hokkaidō Ainu)

	- 'You climbed to [the top of] this mountain.' (Bugaeva 2010, 779)

#### **Set 2.2** (Sakhalin Ainu)


#### **Examples from Other Languages…**

```
Malang Javanese (Austronesian, Indonesia)
(examples from Hemmings 2013, 168-71)
```


*Pelem nyeblòk-i gentèng ómah-ku.* mango fall**-APPL roof** house-1SG.POSS 'A mango fell on the roof of my house.'

*Ibu-ku ng-gupuk kasur nganggó sapu.* mother-1SG.POSS AV-hit mattress AV.use broom 'My mother hit the mattress with a broom (once).'


Malang Javanese has two suffixes (*-i* and *-aké*) that fulfill the primary function of applicatives. Like **prototypical applicatives**, *-i* promotes a noun, that is otherwise expressed via an oblique, to argument of the verb – i.e. applicatives are a **valencychanging strategy**. The applicative *-i* of Malang Javanese typically targets nouns with the **thematic functions** of location or goal. In the first example above you see a non-applicative construction where the location 'on the roof' is expressed via the oblique *menyang gentèng* 'towards the roof'. In contrast, in the second example the verb bears the applicative *-i* and the noun *gentèng* 'roof' is **promoted to argument**. The formerly intransitive verb is now transitive. There are also cases, however, where the applicative suffix *-i* **does not affect** the verb's valency but rather encodes an aspectual meaning (iterative meaning specifically). In such instances, the applicative *-i* has a **valency-preserving** function.

#### 13.3 Analysis and Description

Describe, in no less than 300 words, the semantics and functions of the applicatives *e-*, *ko-*, and *o-* in HA and SA.


**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

### <sup>14</sup> Clause Linking

**Summary** 14.1 Background Information and Observation. – 14.2 Research. – 14.3 Analysis and Description.

#### 14.1 Background Information and Observation

Ainu scholars have traditionally distinguished the clause linking words of Ainu into the two classes of 'conjunctions' and 'conjunctionalisers' (e.g. Refsing 1986; Tamura 2000). In other accounts this distinction is not present and clause linkers are only called 'conjunctions' (e.g. Murasaki 1979). Syntactically, the forms of both classes follow the predicate of the clause they introduce.

Studies that operate this two-fold distinction argue that what separates conjunctions from conjunctionalisers is mainly their prosodic features. On the one hand, conjunctions are said to be free forms as they may be phonologically separated from the predicate of the clause they introduce or may be found 'at the beginning of sentences' (which essentially means after a long pause in speech with a resuming function). On the other hand, conjunctionalisers are said to be bound forms because they form a single phonological unit with the predicate they follow. Nonetheless, prosodic pauses even before conjunctionalisers are sometimes attested, and this behavior eventually obscures the divide with conjunctions. Traditionally, conjunctionalisers have been treated as adverbialisers. As such they are implicitly understood to entail the dependency of the clause they head from a matrix clause (i.e. subordination). However, at present there is still insufficient syntactic evidence to confirm the dependency of

> **Ca' Foscari Japanese Studies 18 Linguistics and Language Education 3 DOI 10.30687/978-88-6969-585-8/014**

clauses headed by conjunctionalisers and, more generally, to investigate the structural properties and differences of conjuctions and conjunctionalisers. Tracing a divide between coordination and subordination in Ainu is a complicated matter and generally conjuctions and conjunctionalisers are implicitly defined as subsuming one or the other on the basis of their contextual uses and translation. In light of these analytical problems and other theoretical ones we will simply talk about 'clause linkers' in this lesson.

Example (1) shows the HA linker *kor* 'while, when' and example (2) shows the SA linker *anah* 'if'. As you can see, in both varieties the clause linker follows the predicate of the clause it heads (put in square brackets in the examples below).

(1) *[… sekor hawean-an] kor a-wen-hoku-hu a-ko-sakayokar.* ADV speak-4S **while** 4-be.bad-husband-POSS 4S-APPL-3SO/insult 'While saying so I insulted my no-good husband.' (Tamura 1985, 30)

(2) *[Kamuy-'ahci i-yee 'an pe nee-no hannehka* god-old.woman 4OI-3SS/3SO/say IPFV NMLZ COP-ADV NEG *'an-kii] 'anah wen.* 4S-3SO/do **if** 3SS/be.bad 'It [would] be bad if I didn't do as the divine old woman tells me.' (Dal Corso 2021)

#### *Consider this additional information…*

In HA and SA the clause linker *kusu* is polysemous as it entails both cause or aim. The causal or final reading is established on context. Examples (3) and (4) illustrate the uses of *kusu* with an example from SA.


#### *Dataset 1 – Semantics of clause linkers*

Look at the following examples that feature the HA clause linkers *wa* and *yak* and the SA clause linkers *teh* and *yahka(yki)*. Consider the translations provided, determine the primary meaning of these linkers, and (if needed) try to propose an equivalent translation that would better fit the context. Then give a closer look at the instances of *wa* and *teh*. Are there cases where the meaning of these linkers is not straightforwardly retrievable in translation? How is clause linking rendered instead? What in the Ainu construction can account for this different interpretation and translation?

#### **Set 1.1** (Hokkaidō Ainu)


#### **Set 1.2** (Sakhalin Ainu)


#### 14.2 Research

Now look at this other dataset. Additional examples from other languages are not given for this lesson.

*Dataset 2 – Semantic overlap*

Examples in the following dataset feature some other clause linkers: *hine*, *hike*, *ciki*, and *yakun* of HA and *koroka*, *waka(yki)*, *ike*, and *wa* of SA. Background context is given in square brackets below each example. Looking at the translation and at the background context try to explain 1) how *hine* and *hike* are different from *wa*, 2) how *ciki* and *yakun* are different from *yak*, 3) how *koroka* and *waka(yki)* are different from *yahka(yki)*, and 4) how *ike* and *wa* are different from *teh*. For each of these linkers the same English translation is given, but you are free to propose a better one according to the context. What kind of semantic relation between clauses do these four groups of linkers entail?

#### **Set 2.1** (Hokkaidō Ainu)


5. *Pewre okkaypoutar ne yakun pokor wa po resu kusu nepki.*

'If they are young men, they have children and they work to provide for them.' (Bugaeva 2004, 249)

[The speaker is imagining how the life of some young men, from whom food has been stolen, must be]

6. *Ienonnoitak yak pirka na!* 'It's good if you pray to me.' (OS 1980, line 480)

[The character is told by a god that he will benefit from praying to him]


'He greeted me and I greeted him.' (Bugaeva 2004, 262) [The character of the story is welcomed by the owner of a house and he responds likewise upon entering]

9. *Arkian yak iociwe kuni eramu.*

'If I come you will think to throw me away.' (OS 1979, line 663) [A dog explains what it expects if it will follow the owner]


'If [it's so], I'll tell it to you.' (Tamura 1984, 18)

[Two people are speaking about a story that A has never heard but about which she is curious]

12. *Aynu okkaypone yaykaran hine pasan kane terkean kane sanan.*

'I turned myself into a human young man and ran downhill leaping.' (Tamura 1985, 60)

[A stoat god describes how he disguised himself in order to then go meet a human young woman who was walking on a mountain trail and deceive her]

13. *Taan hekaci itakan ciki pirkano enu kus ne na.*

'Boy, if I speak you should listen well.' (Bugaeva 2004, 205) [The speaker is about to share important details on his identity with the boy, who is the main character of the story]

#### **Set 2.2** (Sakhalin Ainu)


3. *'Oniste koroka herohki nee nah 'an'eraman kusu sianno keera'anno 'an'ee.* '[The flesh] was firm, but I thought they were *herohki*s so I ate them with much pleasure.' (Dal Corso 2021)

[The speaker is remembering the food she ate at an event and comments on it with a friend]


'Those elders agreed [to the wedding], but the young woman [their daughter] really didn't want to.' (Dal Corso 2021)

[In this tale a young woman refuses to get married and turns down all the suitors she has]

6. *Niicauta 'anse wa sapan.*

'I carried the tree limbs and went [back].' (Dal Corso 2021) [The speaker tells about her activities in the morning]

7. *Neya cih* […] *hesuyehci teh tani 'atuykaene repahci.* '[The whale-gods] went round the boat and eventually went out to the open sea.' (Dal Corso 2021)

[After a father gives away his daughter to a whale god, the group of whales makes its farewell this way before leaving]

8. *Kucasamta sapan teh inuan* […] *ike suke hum annu.* 'I proceeded down to the hut and listened […] and I heard the sound of cooking.' (Pilsudski 1912, 134)

[The character of the tale tells the moments before he met with a she bear that had disguised herself as his wife.]

9. *Ramma ray 'aynu ka 'anhi nee waka sihnu utah tani 'emuyke situsahci.* 'There [had] certainly been people who [had] died, but all the people who [had] survived eventually recovered.' (Dal Corso 2021)

[The characters of the story manage to avoid a famine to get worse by bringing food to people in need, who escape from death thanks to it]

10. *'Anrekucihi kayki* […] *'araka nee koroka kahkemah simma oman kusu* […] *'ariki'anhi nee ko.*

'I also had a sore throat, but because the young woman (= our friend) leaves tomorrow […] I came, you see.' (Dal Corso 2021)

[The speaker stresses how much it meant to her to be present at the last recording session before the collector of the data (i.e. the young woman) returned home]

11. *'Ururukata rikinkehci teh 'ampene 'emus 'ani tatakihci.*

'They lifted [the mouse demon] up onto the river bank and beat it up with the swords.' (Dal Corso 2021)

[After catching the demon that had stolen the village's provisions, a group of man kills it]

12. *Anemakankehe ne numan eperay wakayki ekoyaykus.*

'I sent you [there and] yesterday you fished, but you couldn't [catch anything].' (Pilsudski 1912, 197)

[A god, who had disguised himself as a fish, reveals his plan to the character of the story]

#### 14.3 Analysis and Description

In no less than 300 words give an overview of the semantics of the clause linkers *wa*, *hine*, *hike*, *ciki*, *yak*, and *yakun* in HA and *koroka*, *yahka(yki)*, *waka(yki)*, *ike*, *teh*, and *wa* in SA. Include relevant examples from the sets above where needed.


**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

### <sup>15</sup> Some Uses of Nominalisation

**Summary** 15.1 Background Information and Observation. – 15.2 Research. – 15.3 Analysis and Description.

#### 15.1 Background Information and Observation

In Lesson 5 and Lesson 7 we have seen zero-nominalisation involved in the impersonal-passive and in the negative constructions. Zeronominalisation is just one type of nominalisation present in Ainu. In this lesson we concentrate on the kind of nominalisation that is overtly marked morphologically and on some of its uses in HA and SA.

In HA we find, among others, the nominalisers *-pe* 'thing' and *-(h)i*  'place, moment, fact' (Tamura 2000, 124-7). As for *-pe*, we find the alloform *-p* on vowel-final stems while the form *-pe* is reserved for consonant-final stems. These nominalisers have weak semantics and are specified by the verb/predicate/clause that they follow.

(1) *Mip-ihi ka opitta a-ko-sos-pa-re-hi ka* 3/clothes-POSS even all 4S-APPL-3PO/3POI/3SOI/take.off-CAUS-**NMLZ** even *erampewtek no…* 3SS/3SO/not.know ADV 'Without knowing that I ordered [my magic gloves] to take off [the Water Goddess] all her clothes…' (Bugaeva 2004, 157)

Nominalisation of a verb or predicate may derive a nominal that is re-analysed as a full-fledged noun with its specific semantics. This is called 'lexical nominalisation'.


In SA we find, among others, the nominaliser *-hi* (Dal Corso 2021). The underlying vowel in this nominaliser is *i* but, when *-hi* is attached to vowel-final stems, it changes following the rules of vowel harmony that we have seen for the possessive suffix (cf. Lesson 4).

(4) *Ku-yee-he sunke.* 1SS-3SO/say**-NMLZ** 3SS/be.false 'What I say is a lie.' (Murasaki 1979, 95)

*Consider this additional information…*

Differently from the HA *-pe* and *-(h)i*, the nominaliser *-hi* is not used in lexical nominalisation in SA – in this Ainu variety zero-nominalisation is used to derive nouns like the ones in (2) and (3) above.

(5) *Ipe.* AP.eat(.NMLZ) 'To eat something (> eating something = food).'

#### *Dataset 1 – Types of nominalisation*

Consider the following examples. Some sentences in this set contain the nominalisation strategies we have seen above and others have been included for you to deduce the meaning of words. Note that in Dataset 1 and also in Dataset 2 below nominalisers are given separated from the word they are suffixed to, for more clarity. What is nominalisation used for in these sentences? What interpretations do the nominalised clauses take? Do you think there are any other elements that, together with the nominaliser, derive these specific interpretations of nominalisation as a whole? Do the predicates in the nominalised clauses differ from sentence to sentence in any way (besides in their core semantics)?

#### **Set 1.1** (Hokkaidō Ainu)


'Various [people] had survived, but at that time, like this, a young man came from somewhere carrying a load.' (Tamura 1985, 50)

3. *Aep pirka hi cep pirka hi ikopunpa.*

'They offered me food of the good kind [and] fish of the good kind.' (Tamura 1985, 20)

4. *Somo ka ene yaynuan kunak aramu a p nisapno nea* […] *wenkur eun arpaan rusuy.*

'I thought I wouldn't certainly feel like that, but soon I wanted to go to that […] poor man's [house].' (Tamura 1985, 36)


#### **Set 1.2** (Sakhalin Ainu)

	- 'What I say is a lie.' (Murasaki 1979, 95)

#### 15.2 Research

Now look at this other dataset and also read the examples from other languages given after it, which you will need for the third and last activity of this lesson.

*Dataset 2 – Pragmatic applications of nominalisation*

The following examples illustrate one more use of nominalisation. What is the structural peculiarity of these sentences? How is nominalisation used? Can the nominalised clause in these instances be followed by any other element?

#### **Set 2.1** (Hokkaidō Ainu)

1. *Atuy ruyampe isam yak pirka p!* '[I wish] the sea weren't stormy!' (Tamura 2000, 164)

#### **Set 2.2** (Sakhalin Ainu)


#### **Examples from Other Languages…**

Korean (Koreanic, Korea) (example from Yoon 1991 in Yap, Grunow-Hårsta, Wrona 2011, 6)

*Chelswu-ka chayk-ul ppalli ilk-ess-um-i pwunmyengha-ta.* Chelswu-NOM book-ACC quickly read-PST-NMLZ-NOM evident-DECL 'It is evident (the fact) that Chelswu read the book quickly.'

This example from Korean shows nominalisation of **an entire clause**, that spans from *Chelswuka* to the verb *ilkess-*, marked by the nominaliser *-um*. The verb *ilk-* 'read' retains specification for tense, and it is also marked for nominative case via the postposition *-i*. This last fact indicates that the whole clause is nominalised and functions as a nominal. This type of nominalisation, where the predicate in the nominalised clause is still marked for categories such as tense, aspect, and mood, and the clause is compatible with nominal morphology, is called '**clausal nominalisation**'. A clause nominalised this way functions as a subordinate clause, dependent from a main clause (which, in the example above, is *pwunmyenghata* 'it is evident').

Mongsen Ao (Sino-Tibetan, India)

(example from Coupe 2007 in Yap, Grunow-Hårsta, Wrona 2011, 8)


The example from Mongsen Ao shows an instance of **embedded nominalisation**. Here the coordinated verbs *hwaŋ* 'roast' and *mən* 'sit' in the nominalised clause do not bear markers of tense, aspect, or mood and the whole nominal constituent functions as an argument of the verb in the main clause (i.e. *aɹu* 'be good'). For this reason the nominalised clause is considered embedded (or 'enclosed') in the main predicate. It is important to note that the (im)possibility for the predicate in the nominalised clause to be marked for tense, aspect, or mood **varies from language to language** (Yap, Grunow-Hårsta, Wrona 2011). Therefore, it may be possible to find one or two of these categories marked on the predicate but not the other(s).

Chhantyal (Sino-Tibetan, Nepal) (example from Grunow-Hårsta, Yap 2009 in Yap, Grunow-Hårsta, Wrona 2011, 8)

*Ram-e Sita-o rɦa sat-cyo.* Ram-ERG Sita-GEN goat kill-MIR.NMZ 'Ram killed Sita's goat!' (to the speaker's surprise)

The example from Chhantyal shows a nominalised clause that is not dependent from nor embedded in any main clause or predicate. The nominalised clause, although being morphosyntactically a nominal, functions as an independent clause. This kind of nominalisation is called 'non-embedded nominalisation' or '**insubordination**' (Evans 2007; 2009). Insubordination usually fulfils **specific semantico-pragmatic functions**. For example, in Chhantyal it is used to mark the speaker's surprised attitude towards the event.

#### 15.3 Analysis and Description

In no less than 400 words give an overview of the pragmatic uses of nominalisation you have encountered in Datasets 1 and 2. Cite the relevant examples and also be specific about the kind of nominalisation involved and the morphosyntactic elements that are found together with nominalisation.


**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

# <sup>16</sup> Learning Ainu: Perspectives and Attitudes

**Summary** 16.1 Introduction. – 16.2 Counting Ainu and Ainu Speakers. – 16.3 What is There to Revitalise? – 16.4 The Steps of Language Revitalisation.

#### 16.1 Introduction

After the establishment of the Matsumae Han in Ezo (today's Hokkaidō) in 1604, the Ainu people was subjected to progressive prevarication and oppression by the Japanese. Oppression from the government, social inequity, and the political manipulation of the Ainu identity throughout the following three centuries resulted in permanent damages to Ainu culture. One aspect of the cultural heritage that was particularly affected is language.

Since the foundation of the Matsumae Han and the beginning of Japanese interactions with the Ainu, primarily aimed at establishing business relations, it was necessary to have interpreters to communicate. However, while understanding of the Ainu language was deemed essential for Japanese interpreters, the Ainu were forbidden from learning Japanese to any extent. The Matsumae Clan was in fact afraid that, had the Ainu learnt Japanese, they could have informed the central Tokugawa government of the atrocities that were being committed in Ezo and was therefore interested in maintaining the language barrier (Fukazawa 2019, 7). Establishing such language barrier between the Ainu and the Japanese eventually helped bring the Japanese language to the status of prestige language with respect to Ainu, making this latter the weak language within that newly formed bilingual environment. Therefore, it can be seen how a

> **Ca' Foscari Japanese Studies 18 Linguistics and Language Education 3 DOI 10.30687/978-88-6969-585-8/016**

decline process for the Ainu language was set up right at the start of Ainu-Japanese relations. With time, using the Ainu language was initially discouraged and eventually completely forbidden, at first as a vernacular language to be used in public then more thoroughly even within Ainu communities. The imposition to not speak the Ainu language, kindled by the continuous prejudice that the Ainu people had to face, made Ainu speakers give up their language almost entirely to the point where in many cases it was not passed on to new generations any longer as elders feared negative repercussion on their children and families. This process of language denial was by and large completed by the end of 1920, but some Ainu continued to learn the language in the family, reaching even high levels of proficiency.

#### 16.2 Counting Ainu and Ainu Speakers

This brief introduction to the history of Ainu language in modern times serves to highlight a number of issues that are central when considering the vitality of the language in today's Japan and the efforts towards the revitalisation of Ainu. First, there is the question of how many Ainu are there in contemporary Japan. As Okazaki (2019, 355) points out, the answer to this question is not an easy one and estimates of the number of Ainu vary sharply. Among the reasons for variations in counting Ainu people there is the fact that the polls and censuses carried out to date differ in their scope (for instance, Ainu residing outside of Hokkaidō are never or rarely included). However, one main factor that influences the counting and that easily results in a biased perception of the extension of the Ainu community is that many Ainu people still do not feel comfortable with showing their Ainu identity and therefore do not participate in polls and censuses (Kitahara 2011). As it regards language specifically, there is one more layer of difficulty when it comes to counting speakers of Ainu. Although most native speakers of Ainu have passed away and there are now very few people who acquired the language in the family, there still is a significant number of younger Ainu who can use the language at varying levels of proficiency. This means that the actual number of speakers present in Japan is far more than the alarming figure (5 people) released by UNESCO in 2009. This reality calls for a distinction between 'native' speakers (i.e. those who have learnt the language from a relative and used it as one of their first languages) and 'active' speakers (i.e. those who have a passive and/ or active understanding of the language they have learnt as a second language later in their life and can use it to different extents), which is an aspect of language vitality most important for revitalisation (Okuda 2010).

#### 16.3 What is There to Revitalise?

When it comes to revitalisation there are contrasting opinions even within the Ainu community. Among those who wish for the Ainu culture to be revitalised the majority (53.1%) believes that language is the most important aspect of the Ainu heritage to be preserved for the future. Nevertheless, Ainu ceremonies and dances were also named as cultural heritage that should be given priority. Furthermore, when asked about their Ainu language proficiency and about whether they would consider taking Ainu language lessons, only the 7.2% has said to be able to speak either well or sufficiently well to have a simple conversation and, even more importantly, only the 9.7% showed an interest in learning the language (Fukazawa 2019, 20).

The marginal interest in the Ainu language as a valuable part of the Ainu cultural heritage that could find its place in modern Japanese society has a lot to do with the general public's perspective towards it. As Sawai (1998) notes, the firm convinction that Japan is a monolingual nation and that Ainu is a minority language with not many speakers left has propagated the idea of Ainu as a dying language, too small and with too little space (or no space at all) in contemporary Japan for anyone to spend time for its revitalisation. That is, continuously portraying Ainu as a moribund language through the years has turned into a factual reality and has slowly convinced many members of the Ainu community that any revitalisation effort would be made in vain.

#### 16.4 The Steps of Language Revitalisation

The following excerpts are taken from Kitahara (2012), a paper where the author, as a member of the Ainu community, touches upon a number of issues to be addressed when thinking of revitalising the Ainu language and provides his point of view. Starting from Kitahara's considerations discuss each of the following points.

**1.** Which 'Ainu language' should be revitalised? Though the Saru dialect of Southern Hokkaidō is often tacitly taken as some kind of standard language, Ainu has no real standard variety (Fukazawa 2019, 15-16) and dialectal differences can be striking especially between the Hokkaidō and Sakhalin varieties. Moreover, Ainu has never developed the vocabulary to express concepts and denote things that exist in contemporary society.

*Chiba daigaku un Nakagawa Hiroshi nispa yeehe ene an hi. "Kotan pisno itah katu sinnay ciki, nah wa an kotan un itah neyahka okore nuu easkay pahno wantehci anah pirikahaa. Siisam neyahka, Tokyo kotan un kuru, Osaka un itah kii eaykah yahka, nuu easkay tah nee. Taaha neeno an anah pirika nanko" nah yee. Tani neanpe tah yeeruy pirika anpe nee kuni anramu.* (p. 279)

*Husko ohta isam ike tani aneywanke asiri itah temana anyee kun pe hetaneya. Siisam utah neanpe sianno wooyaan huuresiisam itah nuhci ike siisam itah ne karahci anpe. Nee wahkayki aynu utah neanpe yeyekota an itah kii ruuhe ka isam kusu, siisam neeno asiriitah kara ka hankii. Nee kusu tane aynu itah wante utah an teh itah kii kusu nah eramuokay yahka, husko itah pateh nee anah, anpene itah hayta anpe. Nee teh itah ankara rusuy koroka, itahkara neanpe sianno itah wante utah nee anah easkay koroka, tani sonno hokampa.* (p. 286)

As Nakagawa Hiroshi from Chiba University says: "[Considering] there are dialectal differences, it is good that people learn [the language] enough to understand the variety of whatever village. Even in Japan, a person from Tokyo indeed understands the [Japanese] dialect of Osaka, even if they do not know it. It would be good if [we could] reach this goal". As of today, I believe this would be the best thing.

How is one to express the new words that once did not exist but that are now used? The Japanese heard a lot of words from the Europeans and translated them into their language. However, the Ainu [who] did not even use their mother tongue [could] not create new words like the Japanese. For this reason, even if people now know Ainu and intend to use it, they really lack the terminology [to express themselves] if [they rely] only on the words that already exist. [People have] the will to make up [new] words, but creating neologisms [would] be possible if there were someone who really knew the language, so this is very difficult now.

**2.** What language materials (if any) should be used? How should they be structured?

*Etutaani kotan omoto koro pe nee kusu, yaykotan un itah anecaakasnokara rusuy wakayki, nupurukampi neanpe Saru kotan un itah neewa Chitose kotan un itah pateh koro. Itakirenkakampi neyke, Horobetsu un itah naa koro wakayki, taa itah ahkari an itah neanpe itahhunta ponno ponno pateh anihi nee. Nee kusu asinno anecaakasnokara ohta, Hattori Shirō kara "ainugo hōgen jiten" (1964)* […] *ohta an yaykotan un itah hunara ike PC onne ahunkehci ike imerukorocinunkekampi karahci. Taa pateh neyke, itah hayta kusu, tutanno, wooyaan husko oruspe annuu ike PC onne anahunke ike, opokinno itah anweekaarirehci.* (pp. 280-1)

[The attendants to the courses] were people native of different parts of Hokkaidō, so they wished to learn the dialect of their village, but we had only dictionaries of the Saru and Chitose dialects. As for grammars, there was also one for the Horobetsu dialect but the amount of vocabulary was in fact smaller than that [available] for the other dialects. Therefore, while being taught from the beginning, [the students] looked for the words from their dialect in "A dialect dictionary of Ainu" by Hattori Shirō […], entered them in a computer and created a database. Only by doing this [some] words were [still] missing, so then many old stories were listened to, [words] were included in the PC [database], and [new] terms were added little by little.

**3.** What should be the aim of language revitalisation? How should it be possible to use the language after having learnt it?

*Tah neanpe yeeruye paase ike yeeruye hokampa oruspe nee kuni anramu. Repunmosiri orun oruspe annuu wahkayki, ramma itah ecaakasnokara easkay yahka, nee itah sahno ukoytah easkay pe nee kusu, kii kun pe isam manu. Nee teh itah wante yahka, okaketa neera ka monrayke ne kii ka eaykah anpe yeeruye wen sirihi nee.* (p. 290)

*Nee teh, eh yahka, hoskino anyee pe ani, oya utara ohta yee ka hankii. Tah kusu aynuitah neanpe aynu utuhta pateh kii. Koroka, cise soyta aynu unukara hi anpene ponno pateh an.* (p. 297)

I think this is a more important and more complex matter. Hearing of [analogous] cases from abroad, studying [a language] is always possible but there seems to be no chance to use it because it is in fact possible to communicate without that language. And even if one knows the language, the fact that it cannot be introduced after [one has learnt it] in the workplace seems even worse.

So even if they join [Ainu language courses], as I said before, they do not tell it to other people and for this reason they use the Ainu language only among members of the community. But the chances of meeting [other] Ainu outside of the household remain slim.

**4.** When should one start learning Ainu? Is there a recommended age to ensure proper language acquisition?

*Sianno haciko ohta neanpe henke ahci tura ekihi nee koroka, icaakasnocise ohta ahun ohta otuye. Nee teh yaytuymaaste ike, ramurenkayne sinenehpone neera an pe ka kii easkay pahno poro koh, aynu weekaari ohta oman kuru ka an koroka, pookoro koh poo eyaynuahte kusu suy otuye hemaka. Ene teh, neera an yahka eyaycaakasno kuru neanpe sianno yuhke ramu koro kuru nee ike, porosereke poo reske hemaka teh, monrayki hemaka teh eh.* (p. 297)

[Children] come [to Ainu language courses] at a very young age with [their] grandparents, but when they enter school they drop out [of classes]. Then, once the situation changes and they become old enough to decide for themselves, there are also people who re-join Ainu meetings, but as soon as they have children, they look after them and end up leaving again. That is, a person who [commits to] learning [Ainu] despite [their obligations] is a person with a strong will and, when they are done raising [their] children, the majority of them retires and returns [to meetings and courses].

**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

# <sup>17</sup> The Ainu Language: Bilingualism and Language Education

**Summary** 17.1 Japan – A Monolingual Nation? – 17.2 Bilingual Education in Japan. – 17.3 Teaching and Learning Ainu as a Second Language.

#### 17.1 Japan – A Monolingual Nation?

The notion that Japan is a 'homogeneous', 'monoethnic' and 'monolingual' nation became fixed in the post-war era (Fujita-Round 2019, 172), but in reality Japan has always been at the crossroads of cultural and linguistic exchange through its history. This misconception of Japan as a one-language country, that to much extent still remains in contemporary Japanese society, is mostly rooted in the modern era, namely in the imperialistic period of Japan's history. In the decades leading to the establishment of the Japanese Empire of the 1920s-1930s, Japan annexed four territories: Hokkaidō (1869), the Ogasawara Islands (1872), the Ryukyuan Kingdom (i.e. today's Okinawa province) (1879), Taiwan (1895), and the Korean peninsula (1910). These territories were inhabited by people speaking languages different from Japanese, whom the government made an effort to assimilate among Japanese.

As Morris-Suzuki (1998, 27) notes, imposing Japanese as the national language at all institutional levels during the Meiji period represented the central element of the assimilation process. Especially for the speakers of Ainu and the Ryukyuan languages, who were subjected to assimilation more directly and thoroughly, a forced education imparted only in Japanese slowly resulted in a language shift from the native language to Japanese by the end of the nineteenth

> **Ca' Foscari Japanese Studies 18 Linguistics and Language Education 3 DOI 10.30687/978-88-6969-585-8/017**

century. Importantly, although as a result of this Ainu and Ryukyuan people became in fact bilingual, they were never visibly acknowledged as such. The history of Ainu and Ryukyuan languages illustrates how not recognising bilingualism as a reality has contributed to creating the myth of Japan being a monolingual nation and helps us better understand the difficulties of establishing bilingual education systems in today's Japan.

#### 17.2 Bilingual Education in Japan

Post-war attitudes towards bilingual education in Japan can be seen as a direct result of the assimilation policies of the Meiji period. The case of *kikokushijo* or 'returnee children' provides a good example. Staring from the 1960s, an increasing number children began to arrive in Japan from abroad. These were children of Japanese people who had previously moved abroad for business and whose families were then returning to live in Japan. Having been born and raised in a foreign country, these children obviously behaved and spoke differently from Japanese children who were born and raised in Japan, which took the school system by surprise and for the very first time called for a new approach to education. As it concerned exactly the education of *kikokushijo* children, besides other measures the government gave a series of subsidies for opening special entrance quotas in schools and universities that were aimed at giving support to these children who needed to be re-entered in Japanese society. That is, *kokushijo* children were treated as a minority, in need of public support, who had to be somehow re-Japanised after their long absence from Japan (Fujita-Round 2019, 177-8). Again the reality of bilingualism was essentially denied. Only in the 1970s did the attitude towards bilingual education change and the presence of bilingual people, for whom Japanese may have been either a first or second language, started to be acknowledged more openly. Nevertheless, still today Japanese institutions seem to be slower to adjust to this change of perspective, and in most cases the view that second language learning and bilingualism is of a temporary nature persists. As a reason for this, Kanno (2008) points out on the one hand the teachers' perception of Japanese L2 learners as people who at some point will return to their home countries which somehow allows them to be less invested in their bilingual education. On the other hand, parental attitudes towards children's education are also specifically found to negatively influence the perception of the importance of receiving a bilingual education. Also because of a lack of transparency of the Japanese legislation, parents and children tend to underestimate the value of growing up bilingually and of knowing the language of their country of residency. Other than a substantial change on the Japanese government's part, aimed at addressing bilingualism as a primary aspect of Japanese society, individual attitudes can also make a difference in how a new language can and should be acquired.

#### 17.3 Teaching and Learning Ainu as a Second Language

Read the following quote from Fujita-Round (2019) where the author reasons on some important challenges concerning the future of bilingual education in Japan. In light of these considerations, read the excerpts below taken from Tangiku (2019) who writes about language education for the specific case of Ainu. How should the needs of Ainu speakers and the vitality status of the Ainu language be acknowledged in order to achieve a fruitful revitalisation within an apparently monolingual society?

The difficulty of implementing bilingual education partly comes from the sheer length of time needed to acquire language(s). Moreover, the actual language learning process is individually different, in the context of the society in which the speaker lives. Depending on the speaker's age, bilingual education involves the speaker's language acquisition, language learning, language maintenance and language loss. In some cases, this depends on the position of a language in a society where bilingual education is involved in language endangerment, language death and language revitalisation. Bilingual education cannot be separated from the constant language dynamism of the speaker's life and social reality. Together with the individual difficulty, how to contextualise bilingualism and multilingualism into "bilingual education" will be a challenge for the twenty-first century. (Fujita-Round 2019, 180)

**1.** With the intent of revitalising the Ainu language, educators have adopted some teaching methods from other countries where minority and indigenous languages have been or are being revived successfully. Considering what has been said about perspectives and attitudes of the speakers towards the Ainu language (see Lesson 16), do you think these teaching methods are applicable for Ainu? Are there any aspects, specific to the Ainu case, that should be addressed when drawing from experiences of revitalisation in other countries?

*USA otta ka, Canada otta ka, Hawai'i otta ka, teetawanoankur utar, husko itak oyra okere wa easir, kanna suy kor rusuy utar, yayepakasnu wa tane husko itak eukoysoytak kor oka. Husko itak kanna suy asiknure hi "itaksiknure" "itakmososo" sekor aye p ne na. Husko sinrici kor itak kanna suy yaykata ka konrusuy sekor sanihi utar yaynu. Tane makanak itak asiknure yakun, mososo yakun pirka ya ka aeraman ruwe ne wa oya mositta usa usa husko itak asiknure hawean.* (pp. 166-7)

In the USA, in Canada, and in Hawai'i, indigenous people forgot [their] native language and really [those] people who want to revive it learn it on their own and eventually [can] converse using it. Bringing back the native language is called "language revitalisation" [or] "language reawakening". The descendants of [Ainu] ancestors wish to bring back the language with their own strength. Today it is known how to revitalise [or] reawaken a language effectively and it seems that in different countries native languages are [being] restored.

*Sonno Aynu itak aeaskay rusuy yakun "sinen or wa sinen eun" ani aeraman kuni p ne na. Kanpinuyekur utar neyakka Aynu itak eraman huci utar orowa "sinen or wa sinen eun" ani ayayepakasnu rok pe ne ruwe tapan. Tane oka kanpinuye utar yaykata "sinen or wa sinen eun" ani Aynu itak eraman a korka, pewreutar epakasnu hi ta "sinen or wa sinen eun" ani somo ki no, ramma kane gakko otta neno, Aynu itak eraman rusuy utar sine uske ta uekarpare hi kuoyamokte kor kuan. Kanpinuye utar yaynu hi ene an hi.* […] *"Sinen or wa sinen eun" ani Aynu itak aepakasnu wa tane eraman Yaunkur ka oka. Ponno patek ne yakka oka.* (pp. 168-9)

If one really wants to be able to [speak] Ainu, they should learn it through [the method of] "one-to-one". Even linguists have been studying Ainu from elderly ladies who knew the language with this method. However, today's linguists, [who] have learnt Ainu themselves with the "one-to-one" method, when teaching to younger people do not employ the [same] method [and] always [hold lessons] in schools [where] people who want to learn are gathered [all] in one place. I think this is odd. […] There are also people who have been taught with the "one-to-one" method and now know the language. Though they are a few, there are [some].

#### **2.** As a way to ensure an effective language acquisition, language education is often based on standardised teaching. How can the adoption of the same Ainu teaching materials and methods for everyone (not) satisfy the needs and motivation of individual learners?

*Yaunkur utar usa usa okay kusu utar yaynu hi ka usinnayno an. Husko itak ponno patek eraman rusuy kur ka oka, kestoankor husko itak ani ukoysoytak rusuy kur ka oka. Kes cup an kor sine to ta patek husko itak seysey orowa aepakasnu rusuy sekor an kur ka oka, kestoankor 3 cikan husko itak eukoysoytak rusuy kur ka okay. Usa utar oka kusu ki rusuy pe ka usa kuni p ne na.* (p. 167)

Ainu people are different so their opinions also vary. There are people who want to learn just a little bit of [their] native language, people who want to converse in Ainu every day, and there are also people who say they want to be taught by an Ainu teacher at least one day a month [and] converse in Ainu for three hours a day. Because there are different people, [their] needs must be diverse too.

*Nah an pe neanpe, ikorouncise neya daigaku ohta sicaakasnoyara utah neanpe easkay wahkayki, oya utah neanpe anpene eaykah anpe nee. 2010paa oro AIEA asinno kampisos kara kusu nah yehci ike, paa pisno "nyūmonshū" "shokyūshū" "chūkyūchū" upis ree kampisos, kotan pisno kara kuni urenkarehci. Neewahkayki, etoko wano ankara anpe ka isam ike, anpene kara koyaykus pe neeno aneramuan. Anoka neyke kunine ani kampisos kara aynu anewtanne koroka, kiyanne itah urenkare kuru utah okore montapihci kusu, sine itah urenkare kuru ka sahno ankara kusu karahci. Tah kusu temana ka ankii koyaykus nah aneramuokay kusu, antokoy ne sukuh itah urenkare kuru aneutehkara ike anahunke ike, temana ankara anah pirika nah an pe aneukoramkorohci.* (Kitahara 2012, 281-2)

This method works for people who learn [Ainu during classes] in a museum or at university, but it is really difficult for other people. [Since] 2010 the Organisation for the Revitalisation of Ainu has decided to release new publications [for each Ainu dialect] and, every year, it has published three [volumes] "absolute beginner level", "beginner level", and "intermediate level" [for three different dialects]. However, there was no source already available and this was believed to be an utterly impossible work. I teamed up with the people making the volumes, but older people experienced in Ainu were all busy so it was decided to proceed without a single expert of the language. Because I thought that such project was undoable, I asked a young person I know and included them [in the group] and we discussed how to produce a good publication.

**3.** Besides creating a safe space where Ainu speakers can actively use the language they have learnt in everyday life, revitalisation should also think of finding a place for Ainu speakers to apply their knowledge productively and creatively outside the community and within society. Where to start? Who should be involved in this?

*Aynu itak ne yakka itak ne ruwe tapan. Sisam itak Huresisam itak ka uneno itak ne ruwe tapan. Kamuyyukar aye kusu, yukar aye kusu, inomi aye patek kusu akor itak ka somo ne. Kestoankor tan itak eukoysoytakan kuni p ne ruwe ne. Nep ne yakka aye easkay. Anime (moymoykenoka) otta aye itaki ne yakka aynu itak ani aye easkay.* (p. 170)

*Ne wa oka moymoykenoka ta aynu itak utar, haw kar utar, sisam ne yakka arikikino aynu itak hawehe nu wa, ponno ponno yakka aynu itak ka eraman wa kusu ne no pirkano aynu itak ye hi ruwe tapan.* (p. 171)

Ainu is a language. A language equal to Japanese and Western languages. It is not a language just for reciting the *yukar*s,traditional folklore, and prayers. It is a language to be spoken every day – one can express anything [with it]. Even dialogues in anime can be said in Ainu.

The people who dubbed [the dialogues and] spoke Ainu were Japanese but they worked hard, listened to the language and [eventually] they even understood a little Ainu. This way they in fact [managed to] dub [the anime] in Ainu well.

**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

# <sup>18</sup> Ainu's Relation to Other Languages

**Summary** 18.1 Is Ainu an Isolate Language? – 18.2 Philological Attestations and Investigation. – 18.3 Internal Reconstruction. – 18.4 Comparative Reconstruction. – 18.5 Loanwords and Areal Linguistics. – 18.6 Wörter und Sachen.

#### 18.1 Is Ainu an Isolate Language?

Discuss:


Consider the following six means scholars usually employ to advance their knowledge on isolate languages (Campbell 2017, 11). Do you think they can be applied to Ainu? Which one(s) could be more effective than others and why?


#### 18.2 Philological Attestations and Investigation

#### **List of historical sources on HA:**


#### **List of historical sources on SA:**


For additional sources, also on Kuril Ainu, see Satō, Bugaeva (2019, 69-71).

Attestations of words in historical sources help us trace the path of development of single terms and sometimes of the language in general. Valuable information on adoption of loanwords, creation of toponyms, phonological changes, presently obsolete morphosyntactic structures etc. can surface in historical sources – in the case of Ainu the transcription(s) used represents an insightful tool to investigate language development.

Attestations of *pirka* /piɾi̤ka/ 'be good'; *sirkunne* /ɕiɾi̤kun:e/ 'be dark' (examples from Satō, Bugaeva 2019, 84-6):


The Japanese (*hentaigana*) and Cyrillic transliterations point to actual different phonological realisations for the segment *r* in *pirka* and *sirkunne*. This phonological difference, lost in modern/contemporary Ainu, can be ascribed to different underlying syllable structures for the two words present in Proto-Ainu (see e.g. Satō 2015 for more).

#### 18.3 Internal Reconstruction

Consider again the fourth person prefix *a-* of HA you analysed in Lesson 5 in the function of marker of agent in the impersonal-passive construction. The origin and development of this agreement prefix can be determined through internal reconstructions. Bugaeva (2011, 524-8) traces four stages that led to *a-* being used this way.

The affix *a-* has its origin in the intransitive verb *an* 'exist' in the existential construction, of the type in (1), where this verb takes a noun as its sole argument (S). This is the first, initial, stage.


The second stage of the development sees an intransitive verb being nominalised (via zero-nominalisation, a process widespread in Ainu) and functioning as the S argument of *an*.

(2) *[Rok]S an yakka pirka ya?* **sit.PL.NMLZ** 3SS/exist.PC though 3SS/be.good INT Lit.: 'Is it good if there is sitting down? = May I sit down?' (KJ in Bugaeva 2011, 525)

The verb *an* begins to function as a clitic and semantically the construction soon develops an impersonal reading.

This impersonal construction that was originally limited to intransitive verbs gets extended to transitive verbs by analogy (3). Here the post-clitic *=an* is placed before the verb, thus becoming a prefix, in line with other personal agreement morphemes of Ainu. Erosion of the segment *n*, most common in Ainu when grammaticalisation happens, follows and the prefix takes the form *a-* (the form *an-* still remains in some fossilised verb forms and in North-Eastern Hokkaidō dialects). This third stage of the development is supported by the prosodic features that the personal affix *a-/-an* still retains.


In the fourth and final stage (4) the impersonal construction with a transitive verb takes a passive reading and expression of the actant with an oblique is allowed.


#### 18.4 Comparative Reconstruction

Philological studies of historical sources may include comparative reconstruction and vice versa. Two works that attempt to reconstruct Proto-Ainu using this approach are Vovin (1993) and Alonso de la Fuente (2012).

In particular, Vovin (1993) takes into account the phonetic alternations present in many Hokkaidō and Sakhalin Ainu dialects and, on this basis, sketches a provisional phonemic inventory of Proto-Ainu (PA). Secondly, though far more marginally, Vovin discusses a number of morphophonological variations that occurred in Ainu. Starting from the phonemes of PA some lexical items of the language are reconstructed and then compared with other proto-languages belonging to the Austroasiatic language family (AA).

(PA) *ki=raqu* 'horn'

(PA) \*VqV > (Proto-Wa) +Vŋ (PA) [r] > (PAA, PW, Proto-Monic) [r]

(PA) *suma* 'stone'

(PA) [s] > (PW) [s], (PM) [t], (PAA) [θ]

(PA) *ti=* 'we (personal agreement)'

The comparative reconstruction approach adopted by Vovin is rooted in Hoenigswald's (1960) classical comparative method, with which linguists have long tried to determine cognation among languages. The comparative method includes three steps of language comparison: *step 1* – observe the sound correspondences within words that have a comparable meaning in different languages; *step 2* – infer regular sound changes rules which must explain all sound changes attested in the compared words; *step 3* – infer cognation judgements and propose a phylogenesis for the analysed languages.

The comparative method establishes two types of linguistic characters that help define language cognation: lexical characters (similar words with the same meaning in different languages) and phonological characters (similar or identical sound changes that recur in different languages). A special type of lexical characters are morphological characters (grammatical features such as declension, conjugation, agreement in different languages that can be traced back to the same proto-morpheme). Since morphological material resists borrowing, and therefore it is unlikely that a morpheme in a language changes in an unexpected way, looking at morphology when using the comparative method is most insightful.

In the 1990s, Warnow revises the classical comparative method by stating that, in comparing languages, almost all (but not necessarily all) the linguistic characters should be compatible within the evolutionary tree for the analysed languages. This is a way to bypass the inevitable cases of phonological (e.g. natural sound changes) and lexical (e.g. provable cases of borrowing) idiosyncrasies sometimes attested among languages that otherwise exhibit a strong similarity. As a way to optimise his analysis that, otherwise, would produce several possible trees with varied degrees of compatibility, Warnow formulated the following criterion: *find the tree on which it is possible to explain all incompatible character evolution with as simple an explanation as possible, and which matches linguistic scholarship as closely as possible*. The tree that meets this criterion is the best tree to describe the relations and phylogenesis of the analysed languages.

What are the major pitfalls of this approach?


#### 18.5 Loanwords and Areal Linguistics

Language contact between Ainu and Japanese, on the one hand, and Ainu and Nivkh, on the other hand, is well-substantiated (Dougherty 2019, 101-2). Japanese loanwords in Ainu include *tampaku* (Jap. たばこ *tabako* 'tobacco, cigarette'), Nivkh loanwords in Ainu include *tunakay* (Niv. *tlaŋi* 'reindeer'). The Ainu loanwords that got into Japanese, e.g. ラッコ *rakko* 'sea otter', are rarer.

Grammatical structures or phonological/phonetic features may be shared by languages spoken in a same area (*Sprachbund*). These language features spread among languages through diffusion or contact. One example of diffusion of grammatical structures between HA Ainu and Japanese is e.g. the tentative construction *wa inkar* 'try to', most probably created on calque of the Japanese *-te miru*.

Do loanwords and linguistic features inherited through diffusion and contact tell us something specific about a language genetic affiliation?

#### 18.6 Wörter und Sachen

"Words that can be analysed into component parts are believed to be more recent than words which have no internal analysis" (Campbell 2017, 13). Cases where we can apply this approach in Ainu include common words and many toponyms:

```
oypep (HA), 'oypeh (SA) < 'o-ipe-p (LOC-AP.eat-NMLZ)
'the thing someone eats in = vessel, plate'
aop (HA) < a-o-p (IP-get.in-NMLZ)
'the thing someone gets into = car'
Sapporo < satporo pet (HA) < sat poro pet (be.dry be.big river)
'the big dry river'
```
"Words [that contain non-productive morphology] are assumed to be possibly older than words composed only of productive (regular) morphemes" (Campbell 2017, 14). However, for a language like Ainu…


**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

# Appendix

Transitive agreement paradigms (Lessons 3 and 4).


(Transitive agreement paradigm of HA)


(Transitive agreement paradigm of SA)

**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

# Glossary

#### **Hokkaidō Ainu**







#### **Sakhalin Ainu**<sup>1</sup>


1 In this Sakhalin Ainu Glossary an asterisk is used to signal the underlying form of a word ending in a consonant that is not permitted word-finally (that would be realized as [h]) by the phonotactic rules of the language.






**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

# List of Glosses

The following list summarizes only the glosses used in Ainu language examples. For the glosses used in the examples from other languages please refer to the original work cited. For the sake of consistency and clarity throughout the volume, the glosses for Southern Hokkaidō Ainu examples have been adapted following the conventions employed in Dal Corso (2021). Please refer to Tamura (2000) for the original glosses used by the author, which may differ from the ones reported here.



**Materials and Methods of Analysis for the Study of the Ainu Language** Southern Hokkaidō and Sakhalin Varieties Elia Dal Corso

### Bibliography

Aikhenvald, A.Y. (2004). *Evidentiality*. Oxford: Oxford University Press.

Alonso de la Fuente, J.A. (2012). *The Ainu Languages: Traditional Reconstruction, Eurasian Areal Linguistics, and Diachronic (Holistic) Typology* [PhD dissertation]. Universidad de País Vasco-Euskal Herriko Unibersitatea.

Aoki, H. (1970). *Nez Perce Grammar*. Berkeley; Los Angeles; London: University of California Press. University of California Publications in Linguistics 62.


no-san (a) The Brave Younger Sister of Ishikari Who Foresaw the Night Attack of a Bear). https://ainugo.nam.go.jp/.


### Ca' Foscari Japanese Studies


Per acquistare | To purchase: https://fondazionecafoscari.storeden.com/shop


This volume is intended to be a practical manual to learn the basics of the Ainu language, in its varieties of Southern Hokkaidō and Sakhalin. Thanks to its bottom-up approach and to the activities presented following a growing level of difficulty, this manual is suited for students even superficially trained in general linguistics as well as for the experienced linguist with no previous knowledge of the Ainu language. Through the selected language examples, the reader can also appreciate the regional differences of Ainu and have a glimpse into the Ainu folklore tradition.

**Elia Dal Corso** is a research grant holder and an adjunct professor in Japanese language and Ainu language at Ca' Foscari University of Venice. His research focuses primarily on the Sakhalin variety of Ainu and his specific interests include argument structure, the tense-aspect-mood-evidentiality category, and verb syntax and semantics more generally. He is the author of *The Language and Folklore of West Sakhalin Ainu – A Re-edition of Kyo*̄*ko Murasaki's 'Karafuto Ainugo' with Translation and Grammatical Notes* (Lincom Publishing, 2021).

**18**

**STUDY OF THE AINU LANGUAGE**

DAL CORSO

Università Ca'Foscari Venezia