Degeneracy in historical perspectiveDegeneracy in language • Examples of degeneracy in morphosyntax: –the distinction between main clauses and subordinate clauses by conjunctions

Degeneracythe evolutionary advantage of the violation of

isomorphism

Freek Van de Velde (KU Leuven)

Lauren Fonteyn (University of Manchester)

Introduction

• Constructions: defined as form-function pairings

• Naïve view: this pairing should be fixedIf meaning A corresponds to forms {X, Y, Z}, and form X corresponds to meanings {A, B, C} (many-to-many mapping, instead of Humboldtian isomorphism), then language users have a hard time decoding and encoding language

• Homonymy (polysemy) and synonymy are avoided (Haiman 1980; McMahon 1994: 85)

• However: in reality isomorphism is contstantly violated...

Violations of isomorphism

• Similarity in form begets similarity in function and vice versa (De Smet2010, Fonteyn 2016).

• Superficial (i.e. etymologically unwarranted) similarity may affect the formal realization of neighbouring constructions (Pijpops & Van de Velde2016) and may lead to diachronic merger of distinct lineages (Van de Velde& Van der Horst 2013; Van de Velde et al. 2013)

• Forms with partially overlapping functions may attract each other leading to full overlap in functions (De Smet et al., forthc.).

• In sum, constructions constantly interact on a formal as well as a functional-semantic level. This multitude of complex interactions causes that forms and functions do not exhibit one-to-one, but many-to-manyrelationships.

Degeneracy

• Degeneracy

• Not: the ordinary meaning ('deterioration')

• But: the technical meaning from evolutionary biology: "the ability of elements that are structurally different to perform the same function or yield the same output" (Edelman & Gally 2001:13763)

• Typically, the structurally different elements are at the same time involved in other functions as well.

Examples of degeneracy

• In biology– Thermoregulation in the human body

• shivering

• countercurrent flow

• perspiration

• arteriolar vasodilation

• wearing protective clothing

• huddling

• walking upright

• goose bumps

Degeneracy and related notions

FunctionA

Form1

Isomorphy(1 to 1)

Redundancy(synonymy)

Pleiotropy(homonymy)

Degeneracy(many to many)

Degeneracy is not the same as redundancy, as the strategies have other functions:• Walking upright: increase of visual perimeter, freeing hands ...• Perspiration: removal of toxins• Clothes: fashion• Huddling: release of oxytocin• ...

FunctionA

Form1 Form2

FunctionA

Form1

FunctionB FunctionA

Form1

FunctionB

Form2

Degeneracy in CAS

• Degeneracy is a common feature of so-called 'complex adaptive systems'

• Complex adaptive systems (Holland 1992)– ant colonies (e.g. Hölldobler & Wilson 2008)

– stock market (e.g. Mauboussin 2002)

– human body

– language (Beckner et al. 2009; Bybee 2010; Steels 2011)

• What is the role of degeneracy in CAS?– robustness and evolvability (Whitacre & Bender 2010)

Degeneracy in language

• Language is a complex adaptive system

• Complex adaptive systems display degeneracy

• Languages display degeneracy

– Van de Velde (2014): morphosyntax

– Winter (2014): phonology


• Examples of degeneracy in morphosyntax:

– the distinction between main clauses and subordinate clauses by conjunctions as well as by special word order (as in German Ich weiß nicht, ob er kommenwird (vs. Er wird kommen))

– the marking of the plural by both umlaut and a plural suffix (German Mann –Männer)

– the expression of past time reference both by a prefixed ‘augment’ e- and a suffixed sigmatic marker -s- in Ancient Greek and Old Indic aorists (Ancient Greek é-lu-s-a ‘I unbound’)

• Enhancing robustness and evolvability– in argument realisation by case marking, prepositions, voice distinctions etc.

(see Van de Velde 2014 for diachronic corpus study. See Detges 2009 for similar effects in French)


• Two case studies:– Interrogatives in Germanic

– English deverbal nominalisation

Word order in Germanic

• “Word-position has acquired grammatical significance.” (Jespersen 1993[1894]:111)

• Stepwise development (Faarlund 2001:1708; Hock 2015; Hopper 1975; Van der Horst 2008)

– Proto-Germanic: pragmatically-driven word order with strong tendency to put V in final position

– After desintegration of Germanic unity (> 400-500): rise of V2 - trailblazed by Wackernagel's position of clitic auxiliaries

– As a consequence, the contrast with V1 and V-final was grammaticalized, by contrast:

• V2: declarative main clauses

• V-final (relic): subordinate clauses (backgrounding)

• V1: non-assertional contexts: interrogatives, (irrealis) conditionals, imperatives ... (Daalder 1983; Van der Horst 1984, 1995; Diessel 1997; Goldberg & Del Giudice 2005; Leuschner 2016 (pace Beekhuizen 2016))

Partial degenerate network in Germanic

FunctionA:interrogative

mood

Form2:rising intonation

FunctionB:exclamative

mood

Form1: V1

FunctionC:conditionals

FunctionD:floor holding

Form3:conjunctions

Form4:subject drop

Form5:non-finite forms

FunctionE:temporal

subordination



mood



mood

Form1: V1



Form3:conjunctions

Form4:subject drop


FunctionE:temporal

subordination

diachronically on the rise diachronically on the decline



mood



mood

Form1: V1



Form3:conjunctions

Form4:subject drop


FunctionE:temporal

subordination

insubordination

participle and infinitivalimperatives in Dutch

'uptalk'

Degeneracy in Dutch interrogatives

Sentence type

statement wh-q y/n-q declarative q.

Rising intonation - - +/- +

lexical marker of interrogative mood - + - -

syntactic marker of interrogative mood(inversion)

- (+) + -

Degeneracy in Dutch interrogatives

Sentence type

statement wh-q y/n-q declarative q.

Rising intonation - - +/- +

lexical marker of interrogative mood - + - -

syntactic marker of interrogative mood(inversion)

- (+) + -

Van Heuven (2017)

Degeneracy in (derivational) morphology

• Deverbal nominalization

• many functions?

(e.g. Marchand 1969; Bauer, Lieber & Plag 2013)

event / action object agent

• Deverbal nominalization

• many forms?

(cf. Marchand 1969; Bauer, Lieber & Plag 2013)

Degeneracy in morphology


-ing -ion -ment -al -er



-ing -ion -ment -al -er

eat elect develop approve

complementary distribution > no violation of isomorphism


• Hierarchical nominalization network• Form-function links should be considered on the instantiated level:

Event/action/process E expressed by verb V used as noun NFunction = Refer to event E as participant N using V


• Isomorphism in morphology, a.k.a ‘Blocking’

– ‘the non-occurrence of one form due to the simple existence of another’ (Aronoff 1976: 43)

• Pre-emptive: there can be no synonymy (cf. Bloomfield 1935: 145)

• Diachronic result of competition (cf. Berg 2014): one of pair of synonyms will be lost (cf. Ullmann 1957: 112; Paul 1995 [1920]: 251), e.g. remove vs. removal (Nevalainen 2006)

• Blocking is flawed (e.g. Bauer, Lieber & Plag 2013: 576)

Deverbal nouns

[V + ing]N [V]N [V + ment]N [V + ion]N [V + al]N …

lengthening - - - -

destroying - - destruction -

arriving - - - arrival

shaking shake - - -

moving move movement motion -

Event/action/process E expressed by verb V used as noun NFunction = Refer to event E as participant N using V

e.g. ‘It moved’ > Its move / moving / movement / motion

Deverbal nouns

• Investigating form function links in partial network

• Suffixes taken into account:

• -ing, -(at)ion, -ment, -age, -al, -ance, -ery, Ø (conversion)

• Corpus: CLMET (version 3.1)

• Automatically queried and manually corrected

• e.g. elegance ≠ [eleg]V + ance, passion ≠ [pass]V + ion, train ≠

conversion of [train]V

• After filtering: all suffixes occur with more than one verb stem

1710 - 1780 1780 - 1850 1850 - 1920

tokens 30,000 30,000 30,000

types 1,336 1,657 1,833

Deverbal nouns - network

• But how many of these verbs are shared?• If suffix A and suffix B (form) share a stem V, then A and B are connected through V

• Connections are weighted in terms of how many stems A and B share (weighted

degree)

1710 – 1780

Density: 0.64

Average degree: 72

1780 – 1850

Density: 0.75

Average degree: 111.75

1850 – 1920

Density: 0.72

Average degree: 127.75

Contrastive relations in the network

• There are (somewhat idiosyncratic) contrastive relations between alternating forms on the instantiated/concrete level (cf. Pijpops 2017; Perek 2015: Ch. 5):

– a loud sneeze (one) vs. a loud sneezing (multiple)

– the crack in the ceiling (thing) vs. the cracking of the ceiling (action)

– the edition (thing) vs. the editing (action)

– The record (concrete 💿 ) vs. the recording (abstract 🎶 )

• But do these distinctions always arise because there is a ‘need’ for a one-

to-one organization, or can they be explained otherwise?

Partial degenerate network - pass

• Passage / Passing

• ‘Action related to moving or proceeding on or beyond’ (also die, death)

• (…) Oppose not my passing! (1782)

• (…) an assassin, who waited his passage (1776)

• ‘Approval of a measure, bill, etc. so that it becomes law’

• (…) a brilliant effort to resist the passing of this Act (1820)

• The passage of any measure resembling this would be a (…) (1893)

• Passing place – road – way – corridor

• He had found a passage out to sea (1774)

• O’er [the river] are three passings, and three knight defend … (1872)

• Text

• I shall read a passage of Shakespeare every Sunday (1816)


0

0.5

1

1.5

2

2.5

3

3.5

action - move action - other thing - other road - way text

passage

1710-1780 1780-1850 1850-1920

action thing


0

0.5

1

1.5

2

2.5

3

3.5

4

action - move action - approve action - other road - way text

passing

1710-1780 1780-1850 1850-1920

action thing


action road - way textjourney

passing passage

Degeneracy in nominalization

• In line with CxG theory, the system of English nominalization is considered as a hierarchical network.

• Data indicate that the number of connections (overlapping uses) in the network increases, suggesting that there is no clear tendency towards one-to-one mappings.

• Rather, diachronic development appears to involve a restructuring of many-to-many mappings, in which a form’s loss of meaning can be ‘compensated’ for by strengthening the link between the functional domain involved and another, already available suffix.

• Overlap is not ‘dysfunctional’ in such a system.

Degeneracy - conclusions

• A one-to-one mapping between form and function may not be the gold standard for constructions. There are benefits of a many-to-many mapping (‘degeneracy’): – (1) robustness

– (2) evolvability

• In that sense, language behaves as other complex adaptive systems.

References

• Aronoff, Mark. 1976. Word Formation in Generative Grammar. Cambridge: MIT Press.

• Bauer, Lieber & Ingo Plag. 2013. Oxford Reference Guide to English Morphology. Oxford: OUP.

• Beckner, Clay, Richard Blythe, Joan Bybee, Morten H. Christiansen, William Croft, Nick C. Ellis, John Holland, Jinyun Ke, Diane Larsen-Freeman, Tom Schoenemann. 2009. Language is a Complex Adaptive System: Position Paper. Language Learning 59(S1): 1–26.

• Beekhuizen, Barend. 2016. 'De zijnsstatus van de afhankelijke V１-constructie in het Nederlands'. Nederlandse Taalkunde 21(1): 33-59

• Berg, Thomas. 2014. Competition as a unifying concept for the study of language. The mental Lexicon 9(2), 338–370.

• Bloomfield, Leonard. 1935. Language. Delhi: Motilal Banarsidass.

• Bybee, Joan. 2010. Language, Usage, and Cognition. Cambridge: Cambridge University Press.

• Daalder, Saskia. 1983. 'Aspects of grammatical meaning: the positioning of the Dutch finite verb'. In: R. Jongen, S. de Knop, P. Nelde & M. Quix (eds.), Sprache, Diskurs und Text. Tubingen: Max Niemeyer Verlag, 60-69.

• De Smet, Hendrik, Flach, Susanne & Jukka Tyrkkö. Corpus of Late Modern English Texts (CLMET). Version 3.1.

• De Smet, H., F. D'hoedt, L .Fonteyn & K. Van Goethem. Forthc. ‘The changing functions of competing forms: Attraction and differentiation’.

• Detges, U. 2009. How useful is case morphology? The loss of the Old French two-case system within a theory of preferred argument structure. In: J. Barðdal & S. Chelliah (eds), The role of semantic, pragmatic, and discourse factors in the development of case. Amsterdam: Benjamins. 93-120.

• Diessel, Holger. 1997. 'Verb-first constructions in German'. In: M. Verspoor, K. Lee & E. Sweetser

• (eds.), Lexical and Syntactical Constructions and the Construction of Meaning. Amsterdam:

• John Benjamins, 51-68.

• Edelman, G.M. & J.A. Gally. 2001. ‘Degeneracy and complexity in biological systems’. Proceedings of the National Academy of Sciences 98: 13763-13768.

• Faarlund, J.T. 2001. From ancient Germanic to modern Germanic languages. In: M. Haspelmath, E. König, W. Österreicher & W. Reible (eds.), Language typology and language universals. Vol 2. Berlin: Mouton de Gruyter. 1706-1719.

• Goldberg, Adele and Alex Del Giudice. 2005. Subject-Auxiliary Inversion: a Natural Category’. The Linguistic Review 22: 411-428.

• Holland J. 1992. Complex adaptive systems. Daedalus 121(1): 17-30.• Hopper, P.J. 1975. The syntax of the simple sentence in Proto-Germanic. Den Haag: Mouton.• Jespersen, O. 1993 [1894]. Progress in language. With special reference to English. Amsterdam: John

Benjamins.• Leuschner, Torsten. 2016. 'Afhankelijke v1-zinnen. Conditioneel, contrastief, concessief – functioneel-

typologische en constructionele perspectieven'. Nederlandse Taalkunde 21(1): 61-79.• McMahon, A. M. S. 1994. Understanding Language Change. Cambridge: Cambridge University Press.• Marchand, Hans. 1969. The categories and types of present-day English word-formation: A synchronic-

diachronic approach. München: Verlag.• Nevalainen, Tertu. 2006. Introduction to Early Modern English. • Paul, Hermann. 1920. Principles of Language History. In Auer, Peter & Robert Murray (eds.), Hermann

Paul's 'Principles of Language History' Revisited. Berlin: Mouton.• Perek, Florent. 2015. Argument structure in usage-based construction grammar: experimental and corpus-

based perspectives. Amsterdam: John Benjamins.• Pijpops, D. & F. Van de Velde. 2016. ‘Constructional contamination: How does it work and how do we

measure it?’. Folia Linguistica 50(2): 543-581.• Pijpops, Dirk. 2017. Digging Deeper. The directional alternation in Dutch. BKL, 5 May 2017, Antwerp.• Ullmann, Stephen. 1957. Principles of Semantics. New York: Philosophical Library.• Steels, Luc. 2011. Modeling the Cultural Evolution of Language. Physics of Life Review 8: 339-356.

• Van der Horst, Johannes M. 1984. Over vorm en inhoud van. In: D.M. Bakker (et al.) (eds.). Vorm en funktie in tekst en taal: bundel opstellen verschenen ter gelegenheid van de voltooiing van het honderdste deel van het Tijdschrift voor Nederlandse Taal- en Letterkunde. 154–179. Leiden: Brill.

• Van der Horst, Johannes M. 2008. Geschiedenis van de Nederlandse syntaxis. Leuven: Leuven University Press.

• Van de Velde, F., H. De Smet & L. Ghesquière. 2013. ‘On multiple source constructions in languagechange’. Studies in Language 37(3): 473-489.

• Van de Velde, F. & J. van der Horst. 2013. ‘Homoplasy in diachronic grammar’. Language Sciences 36(1): 66-77.

• Van de Velde, F. 2014. ‘Degeneracy: the maintenance of constructional networks’. In: R. Boogaart, T. Colleman & G. Rutten (eds.), The extending scope of construction grammar. Berlin: De Gruyter. 141-179.

• Van Heuven, Vincent. 2017. 'Prosody and sentence type in Dutch'. Nederlandse Taalkunde 22(1): 3-29.

• Whitacre, James and Axel Bender. 2010. Degeneracy: a Design Principle for Achieving Robustness and Evolvability. Journal of Theoretical Biology 263: 143–153.

• Winter, B. 2014. ‘Spoken language achieves robustness and evolvability by exploiting degeneracy and neutrality’ Bioessays 36: 960-967.

Degeneracy in nominalization

• The network of nominalization strategies scrutinized here is only partial –the system becomes even more intricate if strategies not based on suffixation are included (i.e. ‘downranked clauses’, cf. Halliday 1994).

• Variation in nominalization studies focusses mainly on pairs of alternating forms (that-clause vs. infinitive; infinitive vs. VG; VG vs. deverbal noun: e.g. Dixon 2005; Cuyckens, D’Hoedt & Szmrecsanyi 2014; Brinton 1995; Bloch-Trojnar 2013).

• Distinctions between alternate strategies are probabilistic / not absolute and multifactoral.

• Synthesis of such studies will most likely reveal large degenerate network.

That-clause For…to - inf Emb. quest. Deverbal nounVerbal Gerund

Documents

Degeneracy in historical perspectiveDegeneracy in language • Examples of degeneracy in morphosyntax: –the distinction between main clauses and subordinate clauses by conjunctions