Is grammatical gender less complex than commonly believed? - - PowerPoint PPT Presentation

1

Workshop: “Parallel text analysis in diachronic research” Marburg University, Germany February 22–23, 2018

Is grammatical gender less complex than commonly believed?

Extracting the feminine gender gram from parallel texts

Bernhard Wälchli Stockholm University bernhard@ling.su.se

2

Is gender special or like any other grammatical category? “Gender is the most puzzling of the grammatical categories” (Corbett 1991: 1) Gender is like any other linguistic categories, part of it is simple, part of it is complex

3

You can skip this ad in 5 seconds

Di Garbo, Francesca & Bernhard Wälchli & Bruno Olsson, eds., (forthc.) Grammatical Gender and Linguistic

• Complexity. Language Science Press.

With contributions by Jenny Audring, Östen Dahl, Matthew Dryer, Tom Güldemann & Ines Fiedler, Don Killian, Henrik Liljegren, Matti Miestamo, Johanna Nichols, Kaius Sinnemäki, Erik Svärd, and others.

4

Anaphoric feminine gender gram

5

A major advantage of parallel texts is that language use (to the extent it is not distorted by translationese) can be compared on the level of specific examples Categories can be built bottom-up No need to resort to widely accepted traditional definitions Parallel text typology can hence provide results that are challenging for received views in linguistic theory.

6

3x5 “Base paradigm” of perception verbs (Viberg 2001)

Experience Activity Phenomenon-based SIGHT see look look (like) HEAR hear listen sound ... Ambulatory vision (Mark 5:15): English(leb) and they came to see what it was that had happened. Finnish (1992) Ihmisiä lähti katsomaan, mitä oli tapahtunut.

Wälchli (2017)

• 0.2

0.0 0.2 0.4 0.6

• 0.6
• 0.4
• 0.2

0.0 0.2

Middle English

Dimension 1 Dimension 2 see behold lo see_self look_about look behold about

• 0.2

0.0 0.2 0.4 0.6

• 0.6
• 0.4
• 0.2

0.0 0.2

Finnish (1992)

Dimension 1 Dimension 2 nähdä katsoa katsella huomata pitää_vara katsahtaa varota

see regain sight look around look up look at

• bserve

look watch watch out lo! (go&) see

Ecological psychology (James J. Gibson)

Ambient and ambulatory vision: “One sees the environment not just with the eyes but with the eyes in the head on the shoulders of a body that gets about” (Gibson 1979: 222).

7

Lexical aspect of ‘see’

Following Vendler (1967: 138) linguists quite unanimously ascribe to ‘see’ a dual nature of state (1a) and achievement (1b) (1a) I see Mount Tamalpais. (1b) I reached the crest of the hill and saw Mount Tamalpais. State and achievement profiling of English see according to Croft (2012) a) q seen not seen t b) q seen not seen t

In Russian perfectives of verbs of perception (uvidet’ ‘see[PFV]’) profile the inceptive phase unlike imperfectives (videt’ ‘see[IPFV]) (Croft 2012: 120)

Wälchli (2017)

8

Specificity cline in Bulgakov’s Master i Margarita

joint work with Ruprecht von Waldenfels

• 0.4
• 0.2

0.0 0.2

• 0.2

0.0 0.2 0.4 0.6

Italian

Dimension 1 Dimension 2 vedere [229] scorgere [14] vedersi [6] accorgersi [5] rivedere [3] distinguere [3] scoprire [1] ricognoscere [1]

• sservare [1]

notare [1]

• 0.4
• 0.2

0.0 0.2

• 0.2

0.0 0.2 0.4 0.6

Croatian

Dimension 1 Dimension 2 vidjeti [193] ugledati [60]

• paziti [6]

uviDati [3] spaziti [2] pomotriti [2] pogledati [1] gledati [1]

• 0.4
• 0.2

0.0 0.2

• 0.2

0.0 0.2 0.4 0.6

Russian

Dimension 1 Dimension 2 vidit [129] uvidit [120] razgljadet [13] vidno [4] vidat [4] povidat [2]

• 0.4
• 0.2

0.0 0.2 0.4

• 0.2

0.0 0.2 0.4 0.6

Bulgarian

Dimension 1 Dimension 2 vidja viZdam razliCa zabeleZa razbera zArna razliCavam razbiram nabljudavam liCa

9

Functional domains

“any domain of related (semantic or pragmatic) functions that (one or more) language(s) encode with the formal means they possess” (Miestamo 2005: 293). Functional-domain typology examines the ways in which languages encode functional domains morpho-syntactically or lexically (see Givón 1981; Stassen 1985: 1-23). E.g., comparison: “Definition: a construction in a natural language counts as a comparative construction (and will therefore be taken into account in the typology) if that construction has the semantic function of assigning a graded (i.e. non- identical) position on a predicative scale on two (possibly complex)

• bjects.” (Stassen 1985: 24)

10

Traditional view on gender: formal criteria indispensable “...establishing the existence of a gender system and determining the number of gender requires evidence from agreement (that is, evidence concerned with form). At the same time, gender always has a semantic core: there are no gender systems in which the genders are purely formal categories” (Corbett 1991: 307) “Genders are classes of nouns reflected in the behavior of associated words” (Hockett 1958: 231) Noun classes and agreement as definitional notions Defining notions of agreement according to Corbett (2006: 4): controller, target, (syntactic) domain, features, conditions

• > Gender is complex by definition.

Why not trying a simpler approach and go for the semantic core directly?

11

What exactly is included by noun classes and agreement? (1) English? But he answered: "...." But she said: "..." Japanese? kanojo *‘that.woman’

kanojo wa? she

TOP

‘Do you have a girlfriend?’ (Ishiyama 2008: 232) Kano is originally the attributive form of a distal demonstrative (free form kare) that has come out of use except in a few fixed archaic expressions such as kare kore ‘this and that’. Jo is the Sino-Japanese expression for ‘woman’ (Ishiyama 2008: 141). Kanojo and its masculine counterpart kare ‘he’ (originally ‘that’) were established in the Meiji period (1868-1912) in the literary movement genbun-itchi (unification of written and spoken language) where translations from European languages played an important role (Ishiyama 2008: 139).

12

What exactly is included by noun classes and agreement? (2) Tlacoyalco Popoloca?

=third person pronoun je'e not distinguishing gender =large number of short forms of nouns with anaphoric use (termed “short pronouns” in Stark 2011: 3) (2) Tlacoyalco Popoloca (Stark 2011: 4) Naa janna'a jian anseen jan ixin rinao jan kain

• ne mother

fine heart mother[ANA] because loves mother[ANA] all xe'en jan. children mother[ANA] ‘A mother has a good heart because she loves all her children.’ (Stark 2011: 4) Some condensed anaphoric NPs are reminiscent of noun classifiers (“pronouns that echo a prefix”; Stark 2011: 4) and some uses are compatible with a noun class with agreement interpretation as when animals take the pronoun ba. However, “short pronouns” are productive and apply even to Spanish loanwords (guitaarra, “short pronoun” guitarra).

13

What exactly is included by noun classes and agreement? (3) Iraya (Austronesian, Philippines) person-name markers?

laki Howan (lalaki ‘man’), bayi Mariya (babayi ‘woman’) laki Satanas ‘the Devil’ hadi Dabid ‘King David’ anghil Gabril ‘the angel Gabriel’

Tagalog (Austronesian, Philippines) person-name markers?

si Juan [TOP.PN Juan] ‘John’ ang lalaki [TOP man] ‘the man’

14

Feminine gender as a functional domain

(3) English (Indo-European; Matth. 15:26-27): gender marking on free pronouns But he answered: "...." But she said: "..." (4) Garifuna (Maipurean; Matth. 15:26-27): gender marking on bound pronouns and prepositions Ába l-aríñagun Jesúsu t-un: "...." Ába t-aríñagun: "...." and 3SG.M-say Jesus 3SG.F-to and 3SG.F-say (5) Ama (Arai/Left May; Matth. 15:26-27): gender marking on bound pronouns (S, O) no-na-ni imo na i-so-ki, Isiso mo. that-FOC-here talk

FOC say-O.3SG.F-REM.PST

Jesus

TOP

Ulai no-na-ni nukonu mo na imo-ki, "..." but that-FOC-here woman.SPEC

TOP FOC say[O3SG.M]-REM.PST

(6) Hausa (Afro-Asiatic; Matth. 15:26-27): gender marking on aspect words Ya amsa ya ce: "..." Sai ta ce: "..." 3SG.M answer

PST.3SG.M

say then PST.3SG.F say ‘But he answered: "...." But she said: "..." ’

Functional domain: Anaphoric reference to female human beings

15

 Pronominal gender

Free and bound pronouns: Third person pronouns and affixes for third person have in common that they are reduced referential devices in terms of Kibrik (2011; ch. 3), who calls them free and bound pronouns. (7) Kiribati: (Austronesian, Mirconesian; Matth. 15:27): intermediate referential device Ao e taku neierei ... and 3SG say that[DIST].woman ‘But she said: "..." ’ Gender markers on anaphoric devices can be classified into pronominal (in a wider sense inspired by Kibrik’s terminology) and non-pronominal. Non-pronominal anaphoric devices (intermediate referential devices), such as Kiribati neierei ‘that[DIST].woman’ are less grammaticalized than pronominal gender markers such as English she. Intermediate referential devices tend to be incipient gender markers, nouns on their way to be grammaticalized to pronominal indexes.

16

Intermediate referential device

(8) Yale (Mek, Trans-New Guinea phylum; Heeschen 1992: 29) Nimi ane dinge, bone dinge dane, el-di kwaneng man this property, this.man property

DEM:PL 3SG-GEN sweet.potato

wa-m-la=ba, na do-do de-n. be-PRF-PRS.3SG=CONNECT 1SG take-CVB eat[PFV].PRS.1SG ‘I have taken and eaten (earlier today) this man’s sweet potatoes.’ Table 2: Yale third person pronouns, intermediate forms and demonstrative NPs 3SG N=DEM N DEM el ‘she/he’ bone ‘this.man’ — mene ‘this child’ nimi ane/ene ‘this man’ kel ane/ene ‘this woman’ me ane/ene ‘this child’ For how Nalca (Mek) has acquired gender with noun classes, see Wälchli (forthc.)

17

However, the functional domain approach also includes lexical (non‐grammatical) encoding

(9) Turkish (Matth. 15:24-27) İsa, «...» diy-e cevap ver-di. Jesus say-CVB answer give-PST3 Kadın ise yaklaş-ıp, «...» diyerek [...]. woman however approach-CNV say-CNV İsa

• -na, «...»

dedi. Jesus 3SG-DAT say-PST3 Kadın, «...»

• dedi. «...»

woman say-PST3 ‘But he answered and said, “...” But she came [...] saying, “...”. And he answered and said, “...”. So she said, “...”’

18

Grams and gram types

Gram is a grammatical item in a particular language with specific form and specific meaning and/or function (Bybee & Dahl 1989; Dahl & Wälchli 2016). Grams can be considered in abstraction of the language-specific systems they make part

• f. For instance, perfect and progressive can be investigated in abstraction of tense and

aspect systems. This means for gender grams that the units of research are feminine, masculine, animate and inanimate, rather than gender systems. A gram necessarily has a semantic core, but not all of its uses need be semantically motivated. The gram approach focuses on the semantic core of grammatical categories and investigates to what extent grams across different languages share their semantic core, put differently, cluster to cross-linguistic gram types.

19

Procedural definition of feminine gender gram

(a) Extract all markers picking up reference to female humans (b) unless they can also be used to pick up reference to male humans such as Turkish o ‘he/she’ (c) unless they can also be used to denote female humans in non-anaphoric (notably indefinite) contexts (nouns for ‘woman’, ‘mother’, ‘girl’, ‘daughter’), and such as Turkish kadın ‘(a/the) woman’ (d) unless they are female proper names. such as ‘Mary’

20

Crude practical implementation

a) 76 verses: 40015023 (i.e., Matth. 15:23, Matthew is the 40th book in the Bible), 40023037, 40026012, 41005023, 41005029,

41005042, 41006024, 41006025, 41006028, 41007030, 41010004, 41012021, 41014005, 41014006, 41014008, 42001029, 42001035, 42001036, 42001057, 42001058, 42001061, 42002006, 42002007, 42002036, 42002037, 42002038, 42007013, 42007035, 42007038, 42007047, 42008054, 42008055, 42008056, 42010040, 42010041, 42011031, 42013012, 42015009, 42018005, 42020031, 43004013, 43004016, 43004026, 43008005, 43011005, 43011023, 43011033, 43011033, 43011040, 43012007, 43019027, 43020014, 43020017, 44005008, 44005009, 44005010, 44009037, 44009040, 44012014, 44016015, 44016018, 44016019, 44019027, 45007003, 45009012, 45016002, 46007028, 46007040, 54005006, 54005010, 58011031, 59002025, 66002021, 66002022, 66019008, 66021011.

42015009 (=Luke 15:9) And when she has found it, she calls together her friends and neighbors, saying, ‘Rejoice with me, because I have found the drachma that I had lost!’ 44016015 (=Acts 16:15) And after she was baptized, and her household, she urged us, saying, “If you consider me to be a believer in the Lord, come to my house and stay.” And she prevailed upon us. b) The masculine filter: a form is not extracted if it correlates better with at least one of the following sets: (a) English he, (b)

English him, (c) all uses of anaphoric masculine singular in English together (he, him and his), and (d) all uses of “said to him”

c) The “woman”, “girl” and “mother” filters: a form is not extracted if it correlates better with at least one of

the following sets: (a) the English singular form woman, (b) Xaasongaxango muso ‘woman’ – which is an instance of a very extensive use of a word for ‘woman’, (c) English mother, (d) Nalca gelma ‘girl, daughter’, (e) Upper Pokomo mwanamuke ‘girl’

d) The “Mary” filter e) The “child” filter: a form is not extracted if it correlates better with Tok Pisin pikinini ‘child’ f) The attributive demonstrative filter: removing all forms that collocate better with Folopa kale DEM

21

Results for selected languages

LANGUAGE EXTRACTED FORMS A S P R POSS English (eng) her, she she " her " " Ama (amm) isoki

• V-so- "

V-so- - Garifuna (cab) >#t< t-V "

• , tugía t-un

t-N Hausa (hau) ta, mata, >nta< ta V, ita " V ta mata N-ta LANGUAGE EXTRACTED FORMS REMARKS Esperanto (epo) sxi, >#sxi< sxi-n ACC, sxi-a-AGR POSS Japanese (jpn) kanojo kano-jo PROX-woman Kiribati (gil) neierei, nei neierei F.DIST, Nei female proper name marker Jacaltec (jac) ix ix ‘woman’, noun classifier for ‘woman’ Languages in the sample without any detected anaphoric feminine gender gram: Altaic (10/10); Austronesian (133/134); Niger-Congo (126/127); Quechuan (25/25); Sino- Tibetan (22/24); Trans-New Guinea (79/92); Uralic (7/7)

22

Anaphora and functional domains

(10) English (Indo-European; Matth. 15:26-27): But she said: "..." Saying that (6) reflects the functional domain anaphoric feminine abstracts from the fact that this passage is related to another passage earlier in the text given in (7). In (7), the referent of the anaphor in (6) is introduced in form of an indefinite noun phrase. (11) English (Indo-European; Matth. 15:22): And behold, a Canaanitish woman came out from those borders... Anaphora tend to be co-referent with full noun phrases introduced earlier in the text. However, this does not mean that all anaphora have explicit antecedents with which they are exactly co-referent, as illustrated in (8). (12) Anaphora without explicit antecedent (Hintikka & Kulas 1985: 98): A couple was sitting on a bench. He stood up and she followed his example. The question as to how agreement relates to anaphora is a matter of debate (Croft 2013)

23

Feminine gender gram and noun classifiers

Anaphoric use is a well-recognized function of noun classifiers in some languages. According to Aikhenvald (2000: 87) “noun classifiers are typically used with anaphoric function” (13) Jacaltec (Matth. 14:8) Y-al-ni is-mi' ix t-et tato

E.3-say-DETRANS POSS.3-mother CL.woman/F

3-to COMPL ch-is-k'an ix is-wi' naj Juan;

INCOMPL-E.3-ask CL.woman/F POSS.3-head CL.man/M

John ‘Her mother said that she should ask for John’s head.’

24

Mature (pronominal) feminine gender grams tend to have suppletion and/or neutralization according to grammatical relations

A S P R Poss English she she her her her Belize Kriol shee shee

• German

sie sie ihr ihr ihr-AGR Welsh hi hi hi wrthi ei+ASPIR Latin illa, quæ, hæc illa, quæ, hæc eam, illam

• North. Kurdish

wê

• wê

wê Hindi

• V-ī
• (V-ī)
• Ama
• V-mo-

V-mo- V-mo-

• Au

hɨre / w-V hɨre / w-V V-p V-we AGR-ɨre Mature phenomena imply a lengthy period of historical development and presuppose a non-trivial prehistory (Dahl 2004: 2)

25

Incipient feminine gender grams tend to have invariant forms

Tlalcoyalco Popoloca co jehe xan joanjo xan ngain janné xan and 3 child gave child[ANA] give mother child[ANA]

‘and she (=the girl) brought it to her mother’ (Matth. 14:11)

San Miguel Mixtec te máá‑i, ni ̱ jancha ̱ ka‑i nuu ̱ náa ̱ ‑i and self-YOUNG COMPL gave-YOUNG to mother-YOUNG Tepeuxila Cuicatec ní táⁿ'ā miiⁿ ní ca'a tá chɛɛcu tá and woman.F there/DEF ?

COMPL:give:3 F mother F

San Martín Itunyoso Triqui ni ̱ naga'ui' ún' ra'a nni ún' and gave

F

to mother

F

Chiquihuitlan Mazatec ca-sua na naa rë na

COMPL-give F

mother POSS

F

Amatlan Zapotec nu lee me m-zaaya lo xnaa me and

FOC F COMPL-give to

mother

F

26

Few exceptions Incipient Mature Invariant MANY Malayalam, ... Non‐invariant Yagua MANY

In Yagua, women who have borne children are referred to by dual forms (Payne 1985: 42) – 3DU naada- (often realized as naan-), naadá, 2DU sáána-, saadá – which makes that the pronominal forms for single persons at least for many women is different from that of men – 3SG sa- [I], -níí [II], 2SG jiy- [I], jíy [II]. In the N.T. the second dual forms are used as default for adult women for whom it is not specified in the text whether they have borne children. Even if this is lack of gender from the point of view of the system – and Payne (1985: 42) says explicitly that Yagua lacks gender – this is an anaphoric gender marking opposition from the point of view of language use.

27

Some types of (more or less instructive) mistakes

(a) Anaphoric (demonstrative or definite) forms of a word for ‘woman’: In South Tairora the form nraakyeva [nraakye-va ‘woman-DEM’] is extracted, because the algorithm cannot recognize that it contains nraakye ‘woman’ and should be removed by the “woman”-filter. (b) (Female) proper name markers: A few languages in the sample have female proper name markers but no anaphoric gender. If there are many proper names in the anaphoric domain, the proper name marker can be wrongly extracted: Iraya bayi (shortening of babayi ‘woman’), Uab Meto bi and Satere-Mawe mana. (c) Nominal gender markers: In a few languages female nominal gender markers are

• extracted. This holds both for languages which also have anaphoric gender, such as Abau

sokwe [DIST.DEM.F.OBJ], but also for Parecis -(e)halo and Nalca gera (ge-ra F-TOP). Most nominal gender markers are effectively removed because they correlate better with the “woman”-filter. This holds, for instance, for the Esperanto feminine noun suffix -ino. (d) Masculine gender for female speakers: While the quotations do not do any harm for most languages, in Kayabi (Tupian) they cause the error that kĩã ‘M 3SG (female speaker)’ is wrongly extracted.

28

Translationese As there are sometimes attempts to eliminate anaphoric gender by language planning (for instance, in Swedish a gender neutral form hen has been suggested to replace han ‘he’ and hon ‘she’), there are some attempts to implement gender distinctions in pronouns where there a none. A case in point is Uduk where the N.T. uses the noun (a)yim [CL2] ‘female friend’ for ‘she’ even though this noun does not have any anaphoric use in spoken Uduk (Don Killian, p.c.). Thus Bible translation Uduk has a special pronominal noun whereas there are no indications of a grammaticalization

• f an anaphoric gender gram in spoken Uduk (for more information on

gender in Uduk see Killian, forthc).

29

Where the algorithm fails to detect gender

(a) There is agreement gender, but no anaphoric gender: Limbu, Iraqw, Baruya, Biangai, Kadiweu. (b) Gender is distinguished in pronouns, but only in the second or in the second and first persons: Iraqw, Basque, Paez. (c) There is anaphoric feminine gender, but only for girls, the adult women domain is covered by a general human respect gender: Peñoles Mixtec Coatzospan Mixtec, Texmelucan Zapotec. These are removed by the “girl” filter. (d) Gender marking is restricted to a limited part of the S and P domain and the markers do not have high cue validity: Chechen, Hindi, Gujarati, and Eastern Panjabi. The anaphoric function in those languages is marginal. (e) The marker is partly zero as opposed to a non-zero masculine marker: Ashaninka, Caquinte, Pichis Asheninka and Nomatsiguenga. The algorithm as implemented here is not smart enough that it can recognize zero as the marker of the feminine gender gram. (f) The dominant marker is orthographically identical with another form: Teutila Cuicatec.

30

Anaphoric gender has high cue validity

In most languages anaphoric gender markers can be identified without previous analysis of any other grammatical categories or lexemes. This means that some anaphoric gender markers tend to have high cue validity in most languages and are constructional islands (item-based constructions with a constant element; Tomasello 2003) which can be considered in abstraction from most other aspects

• f grammar as a form-meaning relationship in the text.

Cases of “diagonal” syncretism Latvian Afrikaans

F M NOM.SG viņa viņš GEN.SG viņas viņa F M

3SG sy hy

POSS.3SG haar sy

All languages with “diagonal” syncretism in the sample also have another anaphoric feminine gender marker with higher cue validity.

31

Proficiency mode

Native speaker, Language Expert Typo log ist ?

Discovery mode

L1 learner, L2 learner

CONSTRUCTION

syntactic properties morphological properties phonological properties semantic properties pragmatic properties discourse-functional properties

FORM

symbolic correspondence (link)

(CONVENTIONAL)

MEANING

syntactic properties morphological properties phonological properties semantic properties pragmatic properties discourse-functional properties

FORM MEANING

32

Anaphoric feminine gender gram

33

Conclusions =Parallel texts help shift the focus of attention to the most functional aspects of gender which makes gender appear as a much less exclusive category type than traditionally believed =Part of gender behaves like any other grammatical category with a semantic core. No need to resort to such complex notions as noun classes, agreement and system. =Gender with agreement, noun classes, in systems, with gender marked pronouns etc. is mature gender. Restricting gender by definition to mature gender is not profitable for the study of the origin

• f gender.

=Suppletion and/or neutralization is characteristic of mature gender (diachronically more stable)

34

References (selected):

Aikhenvald, Alexandra Y. 2000. Classifiers. A Typology of Noun Categorization Devices. Oxford: Oxford University Press. Bybee, Joan & Dahl, Östen. 1989. The Creation of tense and aspect systems in the languages of the world. Studies in Language 13.51–103. Corbett, Greville G. 1991. Gender. Cambridge: Cambridge University Press. Corbett, Greville G. 2006. Agreement. Cambridge: Cambridge University Press. Croft, William. 2012. Verbs. Aspect and Causal Structure. Oxford: Oxford University Press. Croft, William. 2013. Agreement as anaphora, anaphora as coreference. In Dik Bakker & Martin Haspelmath (eds.), Languages across Boundaries: Studies in Memory of Anna Siewierska, 107–129. Berlin: Mouton de Gruyter. Dahl, Östen & Bernhard Wälchli. 2016. Perfects and iamitives: two gram types in one grammatical space. Letras de Hoje 51: 325-348. Dahl, Östen. 2004. The Growth and Maintenance of Linguistic Complexity. Amsterdam: Benjamins. Gibson, James J. 1979. The Ecological Approach to Visual Perception. Reprinted 2015 Givón, Talmy. 1981. Typology and functional domains. Studies in Language 5: 163–193. Heeschen, Volker. 1992. A Dictionary of the Yale (Kosarek) Language (with Sketch Grammar and English Index). Berlin: Reimer.

35

Hintikka, Jaakko & Kulas, Jack. 1985. Anaphora and Definite Descriptions. Two Applications of Game-Theoretical Semantics. Dordrecht: Reidel. Hockett, C. F. 1958. A Course in Modern Linguistics. New York: Macmillan. Ishiyama, Osamu. 2008. Diachronic perspectives on personal pronouns in Japanese. Ph.D.

• thesis. State University of New York at Buffalo.

Kibrik, Andrej A. 2011. Reference in Discourse. Oxford: Oxford University Press. Miestamo, Matti. 2007. Symmetric and asymmetric encoding of functional domains, with remarks on typological markedness. In Miestamo, Matti & Wälchli, Bernhard (eds.), New Challenges in Typology: Broadening the Horizons and Redefining the Foundations, 293-314. Berlin: Mouton de Gruyter. Payne, Doris Lander. 1985. Aspects of the Grammar of Yagua: A Typological Perspective (Peru). University of California, Los Angeles doctoral dissertation. Stark, Sharon L. 2011. Ngigua (Popoloca) pronouns. SIL-Mexico Branch Electronic Working Papers #012. http://www.sil.org/mexico/workpapers/WP012i- PopolocaPronouns-pls.pdf Stassen, Leon. 1985. Comparison and Universal Grammar. Oxford: Blackwell Tomasello, Michael. 2003. Constructing a Language: A Usage-Based Theory of Language

• Acquisition. Harvard University Press.

Vendler, Zeno. 1967. Linguistics in Philosophy. Ithaca, N.Y.: Cornell University Press.

36

Viberg, Åke. 1984. The verbs of perception: a typological study. Linguistics 21.1, 123– 162. Viberg, Åke. 2001. Verbs of perception. In: Martin Haspelmath, Ekkehard König, Wulf Oesterreicher & Wolfgang Raible, eds., Language Typology and Language Universals: An International Handbook. Berlin: Walter de Gruyter, 1294–1309. von Waldenfels, Ruprecht. 2006. Compiling a parallel corpus of Slavic languages. Text strategies, tools and the question of lemmatization in alignment. In: B. Brehmer, V. Zdanova & R. Zimny, Hrsg., Beiträge der Europäischen Slavistischen Linguistik (POLYSLAV) 9. München, 123–138. Wälchli, Bernhard & Cysouw, Michael. 2012. Lexical typology through similarity semantics: Toward a semantic map of motion verbs. Linguistics 50.3: 671-710. Wälchli, Bernhard. 2017. Non-specific, specific and obscured perception verbs in Baltic

• languages. Baltic Linguistics 7.

Wälchli, Bernhard. Forthc. The rise of gender in Nalca (Mek, Tanah Papua): the drift towards the canonical gender attractor. In Audring, J., Fedden, S. & Corbett, G. (eds.) Non-canonical gender systems.