Grammaticization in Language Evolution

1
GrammaticizationinLanguage Evolution

• Salikoko S. Mufwene
• University of Chicago

2
Grammaticization as a diachronic process

• Grammaticization as historical change
• comparable to semantic shift
• It is a functional shift
• It is gradual a) in the way it evolves b) in
  the way it spreads within the community
• It is part of language evolution
• It is defined by its outcome, not as a specific
  kind of restructuring process
• It is unidirectional, but not necessarily
  unilinear nor rectilinear

3
Grammaticization as semantic shift

• Chair N, furniture ? chair N, position of
  authority ? chair V, preside over
• (up)on NP ? (up)on V-ing ? a-V-ing ? V-ing
• on board/shore ? aboard/ashore
• Creoles a V (as in im a kom)
• The evolution coopts extant morphosyntactic
  principles, such as predication with a copula
  when the head of the predicate phrase is
  non-verbal in English, whereas no such constraint
  applies in creoles
• The creole pronunciation of a reflects its
  phonology

4
Grammaticization is gradual

• the preposition must combine with a nominal
  object in English
• The bleaching of the preposition causes a
  reanalysis of the verbal noun
• Creoles differ from modern standard English
  partly in regard to what carries the PROGRESSIVE
  meaning (Cultural aspect of evolution)
• Variation in creoles reflects the extent to which
  the construction has spread among speakers

5
Grammaticization is part of language evolution

• It contributed to the gradual divergence of
  creoles from their lexifiers
• It is also evident in go(ing) to N ? going to V
  ? gonna V ? gon V
• It is also a process of idiomatization (gonna N)
• Creoles a) a go V b) go V
• No break in the evolution from the lexifier to
  creoles, though the influences are multiple
• Multiple influences conjures up variation,
  compe-tition, and selection in this particular
  evolution
• Reanalysis cannot be separated from
  grammaticiz-ation

6
Grammaticization is a functional shift

• Lexical category shift down-N (hill) ?
  down-P/ADV
• ... often concurrent with reanalysis by side
  (PN) ? beside by hind ? behind by fore ?
  before
• Do composite prepositions such as in front of, in
  back of, and in addition to represent
  transitional stages?
• reminiscent of quantifier phrases such as a
  number/amount of (Note also a certain amount of
  an unspecified number of)
• A basic mechanism in grammaticization is the
  cooptation of extant items or structures for new
  functions, just like exaptation in biological
  evolution

7
Etymologies are informative

• The processes/strategies may originate in the
  proto-language, as in the case of FUTURE and
  PERFECT constructions in French
• They may originate in another language, such as
  the article el in Spanish
• Invocations of apparent grammaticalization in
  creolistics are unjustified
• Language speciation is a consequence of
  population movements and language contacts, which
  produce alternative strategies or new ways of
  using extant materials

8
Polygrammaticalization as bifurcated evolution

• Evidenced by layering New patterns do not
  necessarily displace the older ones, e.g. the
  gram-maticization of side, back, front, etc.
• Two or more processes of grammaticization can
  occur concurrently, as with fu/fi/f? lt Eng for
  f? in Atlantic English creoles Complementizer
  fu as in Im wan fu go and the purposive modal fu
  as in Im (ben) fu go.
• This is an evolution facilitated by the for-to
  comple-mentizer construction in English and by
  the option of verbless predication in creoles and
  some substrate languages
• Two creoles can also take divergent paths, as
  with a go V in Jamaican Creole but go V in
  Gullah

9
Grammaticization often contributes to (more)
variation and therefore to competition

• Alternative FUTURE constructions in Jamaican
  Creole wi V vs a go V
• Specialization of FUTURE constructions in
  standard English will V vs be going to V
• Inter- and intra-idiolectal variation, e.g.,
  Gullah DURA-TIVE V-in vs d? V vs d? V-in
• Time may (not) resolve the competition, though
  selection need not eliminate variation

10
Universal pathways of grammaticization?

• Universal tendencies and alternatives, e.g.,
  cooptation of terms for body parts to denote
  location and of locative expressions for temporal
  reference (e.g., she stood/came behind me) the
  cooptation of want, go, or come for FUTURE
  constructions
• Largely a function of extant strategies in the
  language (variety), as in (in) front/back (of)
  me However, in Sranan (na) baka a oso vs oso
  baka

11
The meaning of emergent grammar

• Speakers as makers of their languages, during
  their communicative acts
• through their innovations and deviations
• At the idiolectal level, repetition and extension
  of the new behavior to related items restructure
  the system
• At the communal level, the restructuring that
  matters is what is copied by and spreads among
  other speakers
• Emergent grammar does not deny the existence
  nor significance of grammar as generalizations
  over the behaviors of classes of items

12
There is no restructuring process that is unique
to grammaticization

• Not semantic/functional shift
• Not reanalysis
• Not semantic bleaching and generalization
• Not the idiomatization of the relevant
  construction
• Not the phonetic reduction
• Only the outcome makes grammaticization
  different the specialization of a
  form/construction for a grammatical function

13
The debate on unidirectionality seems misguided

• Unlike movement in space, evolution proceeds in
  time and in one direction
• although it can be bifurcated, multilinear
• It need not be rectilinear
• No evidence of degrammaticization has been
  adduced that illustrates a reversal of the
  grammati-cization trajectory e.g., if down-V has
  evolved from down-P, the latter did not evolve
  from the former.

14
Hints about the evolution of language

• The structural and function exaptations
  observable in grammaticization are not unique to
  this process
• Most such exaptations must have occurred several
  times in the hominid phylogeny, coopting
  different components of the extant communicative
  means
• Such exaptations are not planned and are
  initiated by individuals, from whom they spread
  within the wider population
• Different individuals do not innovate in
  identical ways, a state of affairs which fosters
  variation, competition, and selection
• The hidden hand produces the communal norm,
  largely through mutual accommodations among
  speakers

15
Hints about the evolution of language - 2

• Social-interactional dynamics make all changes
  gradual at the communal level
• Ecological changes and ontological mutations
  within the hominid line militate concurrently for
  a gradual perspective on the evolution of
  language
• There is not much evidence for punctuated
  equilibrium or catastrophic evolution
• though the speed of evolution has not been the
  same at all stages of the hominid phylogeny

16
Hints about the evolution of language -3

• The evolution of language seems to have proceeded
  more or less like that of computers, with the
  inter-vals of innovations becoming shorter and
  shorter and their consequences ranging wider and
  wider later and later in time.
• But I really need a separate paper on different
  stages of the evolution of language to support
  these speculations with paleontological
  evidence.

17
Thank you!http//humanities.uchicago.edu/facult
y/mufwene/then click on Goodies for some
heresies