Unit 10: Language I
questions to ask about language
- (I) What language is FOR
- (II) What language is LIKE
- (III) What it means TO KNOW language
- (IV) How language is LEARNED and USED
(I) what language is [NOT] for: conveying "meaning"
The standard [formalist / functionalist] theory of what language is
for: structured thinking / communication, as in —
"Take a meaning, encode it, place it in an envelope, give
it to the other side, which then decodes it".
On that theory, here's the conception of "meaning", part I:
lexical semantics, or the meaning of "words" —
Trotsky saw a hyrax
Trotsky saw a capybara
Trotsky saw a marmot
An alternative conception: "words" as the handles of TOOLS.
(I) what language is [NOT] for: conveying "meaning"
The standard conception of "meaning", part II:
compositional semantics —
The meaning of a "sentence" is a function of the meanings of its
parts.
An alternative conception: uttering sequences of "words" = working with
TOOLS.
(I) what language is FOR (after Jakobson, 1960)
Jakobson's
six functions of language:
- emotive
- poetic
- phatic
- referential
- metalingual
- CONATIVE
(II) what language is LIKE: it's like a lot of other behaviors
"... the coordination of leg movements in insects, the song of birds, the
control of trotting and pacing in a gaited horse, the rat running the maze,
the architect designing a house, the carpenter sawing a board present a
problem of sequences of action ..."
— Karl Lashley (1951)
A thing to remember: when someone insists on the uniqueness of human
language, ask about individual differences and also about any animal
communication systems whose study may be relevant. If they tell you that
none are, persist.
Another thing to remember: when someone describes to you a study in psychology
or neurobiology of language, ask about individual differences.
(II) what language ABILITY is like: individual differences in linguistic ability
The LAD goes to school: A cautionary tale
for nativists
Ewa Dąbrowska, Linguistics 35:735-766 (1997).
Four types of sentences: complex NP,
tough
movement,
parasitic gap (two types), and
controls; 20 sentences altogether. Examples:
Q1. Robert was convinced that Jerry would get the job even though he
wasn't really qualified for it. [control]
What was Robert convinced of?
What wasn't Jerry qualified for?
Q2. Sandy will be easy to get the president to vote for.
Who will vote?
For whom will this person vote?
What will be easy?
Who will find it easy to do this?
(II) what language ABILITY is like: individual differences (cont.)
The LAD goes to school: A cautionary tale for nativists
Ewa Dąbrowska, Linguistics 35:735-766 (1997).
Four types of sentences: complex NP, tough movement, gap 1, gap 2, and
controls; 20 sentences altogether. More examples:
Q3. It was King Louis who the general convinced that this slave might
speak to.
Who might the slave speak to?
Who did the convincing?
Who was convinced of something?
Q5. The manager knew that the fact that taking good care of herself was essential upset
Alice.
What did the manager know?
What upset Alice?
(II) what language ABILITY is like: individual differences (cont.)
Test scores, by subject population type (max possible score was 640):
The LAD goes to school: A cautionary tale for nativists
Ewa Dąbrowska, Linguistics 35:735-766 (1997).
(II) what language ABILITY is like: individual differences (cont.)
Test scores, by sentence type and subject population type:
The LAD goes to school: A cautionary tale for nativists
Ewa Dąbrowska, Linguistics 35:735-766 (1997).
(II) what language is like: two complementary computational aspects
The problem of serial order in behavior, as formulated by Lashley
(1951), contains, upon reflection, two complementary aspects:
- The STRUCTURAL or SYNCHRONIC aspect: how to deal with
multiple input and/or output variables, specified at a given
instant of time. Hence the computational problem
of multimodality: embodied agents must
deal with multiple streams of information (multiple sensors &
actuators), which, moreover, may differ radically in their
statistical and other properties (e.g., auditory vs. visual
cues).
-
The TEMPORAL or DIACHRONIC aspect: how to deal with multiple
streams of information as they unfold over time. Hence the
computational problem of
concurrency: the dynamical processes that affect / comprise
behavior not only unfold in parallel: they generally do so at
different rates and independently, or asynchronously, with
regard to each other.
[O. Kolodny and S. Edelman (2015).
The problem of multimodal
concurrent serial order in behavior, Neuroscience and
Biobehavioral Reviews 56:252-265.]
(II) what language USE is like: the dialogic structure of language
[John Du Bois
(2014). Towards a dialogic syntax, Cognitive
Linguistics 25:359-410]
(II) what language USE is like: the dialogic structure of language
[John Du Bois
(2014). Towards a dialogic syntax, Cognitive
Linguistics 25:359-410]
(II) what language USE is like: the dialogic structure of language
Above: partially ALIGNED and MATCHED halves of a snippet of
dialogue from the Santa Barbara Corpus of Spoken American English.
Right: partially ALIGNED and MATCHED portions of a
GRAPH
representation of the (joint) corpus.
Graphs are a useful computational tool for representing
discrete sequential (and hierarchical) structure.
(II) what [A RECORD OF] language USE is like
[...] The grammatical possibilities of a language are like a
network of paths in open grassland. There are a number of heavily
used, wide and well-beaten tracks. Other, less popular routes are
narrower, and the variation extends smoothly down to routes used only
very occasionally, which are barely distinguishable furrows or, if
they are used rarely enough, perhaps not even visible as permanent
marks in the grass; but there are no fences anywhere preventing any
particular route being used, and there is no sharp discontinuity akin
to the contrast between metalled roads and foot-made paths.
— G. Sampson (2007)
(II) what [A RECORD OF] language USE is like: a GRAPH
Think of words as stations, sentences as using the
system for travel (possibly changing trains).
(II) what [THE DYNAMICS OF] language USE is like: production and dialogue
(III) what it means TO KNOW language [the standard story]
On the standard [formalist / functionalist] account, to know language means
to possess a GRAMMAR — a system of formal rules that (i) generate (all
and only well-formed) sentences and that (ii) support parsing of
sentences into their constituents.
An alternative theory that is gaining ascendance posits
a LEXICON/GRAMMAR that is a collection of patterns
(constructions) that are distilled from experience
through statistical inference.
These patterns are generative: they can give rise to partially new
utterances and behaviors that further enrich the communal experience.
(III) what it means to know language
The radical (and, I think, correct) view is that constructions form
a TOOLKIT for INFLUENCING other people's thinking and behavior (as
well as one's own).
To know language means to have mastery of a toolkit that allows
one to influence behavior and thinking, of others and of self.
(III) what it means to know language: the core computational problem
The general computational problem implied by the above
characteristics is that of control. The need for control is
universal in the animal kingdom; as Savage-Rumbaugh and Lewin (1994)
put it: “All organisms with complex nervous systems are faced with
the moment-by-moment question that is posed by life: what shall I do
next?”
Language behaviors reside in a graph-like space of sequences of
multimodal gestures that embody a game of influence between
interlocutors and that are subject to multiple, dynamically changing
influences and contextual constraints. THIS IS THE SPACE THAT NEEDS
TO BE LEARNED.
The ultimate problem of control in language is, however, the same as
in any other behavior: whereas its general formulation is “What
shall I do next?”, in language it becomes “What shall I say [and do]
next?”. Thus, insofar as it requires choosing a course of action,
behavioral control is always an exercise in navigation: “Where, in
the space of possible choices, should I go?”
