book excerptise:   a book unexamined is wasting trees

Psychology of language

David W. Carroll

Carroll, David W.;

Psychology of language

Cengage Learning, 2008, 492 pages

ISBN 0495099694, 9780495099697

topics: |  language | psychology

1 Introduction


Generally speaking, much of our linguistic knowledge is tacit
rather than explicit. p.4
Tacit knowledge refers to the knowledge of how to perform various
acts, whereas explicit knowledge refers to the knowledge of the processes or
mechanisms used in these acts.

[Opens with four examples.  Each underlines an aspect of language
processing. ]
a) Gardenpath Sentence: "He accepted the deal before checking his finances.
   He was in a quandary when he saw he had a straight flush."
   "deal" : financial deal -> cards
b) Indirect request : "Can you open the door?" [vs. "Open the door!"]
   why are indirect requests more "polite"?  --> SOCIOLINGUISTICS
c) Aphasia: (Wernicke's): Before I was in the one here, I was over in the
	other one. My sister had the department in the other one. (Geschwind,
	1972, p. 78)
   displays appropriate syntactic structure and phonologically,
   was articulated smoothly and with appropriate pausing and intonation.
   But semantic relationships are seriously disrupted --> NEUROLINGUISTICS
d) Language in children:
   1-yr-o child is struggling with her shoes and
   the mother asks her what she is doing, the simple response is "off".

   although there is disagreement over exactly how much knowledge to
   attribute to young children, it appears that children know more than they
   say.
   somewhat older children: "baby gone" - eliminates closed-class or function
   words (prepositions, conjunctions, and so on) in favor of open-class or
   content words (nouns, verbs, adjectives).
    This pattern suggests that children have an intuitive understanding of
    these two grammatical classes, which is part of their syntactic
    knowledge. p. 8  --> DEVELOPMENTAL

History of psycholinguistics




the interdisciplinary field of psycholinguistics flourished twice: 
 1. around the turn of the last century, principally in Europe, and
	linguists turned to psychologists for insights into how human beings
	use language.
 2. in the middle of the 20th century, principally in the United States.
	Now psychologists turned to linguists for insights into the nature of
	language. 

[observation from
  Blumenthal, A. L. (1987). The emergence of psycholinguistics. Synthese, 72,
  313-324.]

In between these two periods, behaviorism dominated both fields, each of
which practiced a form of benign neglect toward one another.

[?? self-contradicts later in assessing Bloomfield as a behaiourally oriented
linguist.  Perhaps it would be more useful to characterize four, or even five
periods:
a) Cartesianism: the idea of mental, conscious language; Port Royal
	grammarians - 1600-1900
b) Wundt / James - systematization of mental processes - 1880-1930s
c) Behaviourism - rejection of mental processes 1915-1960
d) Chomskianism - rejection of semantics - 1957-2000
e) Cognitive Science - integration of views from psychology, neuroscience,
   computation, linguistics. 1980-now
]

Introspective approaches

Wundt's theory of language production: regarded the sentence, not the word,
   as the primary unit of language and saw the production of speech as the
   transformation of a complete thought process into sequentially organized
   speech segments (comprehension was thought to be basically the same
   process in reverse). describes speech production:

	When I construct a sentence, an isolated concept does not first enter
	consciousness causing me to utter a sound to represent it. That it
	cannot be this way is shown by the phenomenon of phonetic induction
	which occurs when a vocal element on the verge of being expressed is
	already affecting the form of a sound being spoken at the moment.
	And similarly, an articulation that has just occurred influences the
	succeeding sound. . . . The sentence . . . is not an image running
	with precision through consciousness where each single word or single
	sound appears only momentarily while the preceding and following
	elements are lost from consciousness. Rather, it stands as a whole at
	the cognitive level while it is being spoken. (Wundt, 1912, cited in
	Blumenthal, 1970, p. 21)

From the 1920s to the 1950s, psychologists expressed relatively little interest
in language. Behaviorists preferred instead to speak of ‘‘verbal behavior.’’ The
nbehavior of speaking correctly was, it was assumed, the consequence of being
raised in an environment in which correct language models were present and
in which children’s speech errors were corrected. The manner in which parents
shape their children’s utterances was described by the behaviorist B. F. Skinner
(1957) in his book Verbal Behavior:

    In teaching the young child to talk, the formal specifications upon which
    reinforcement is contingent are at first greatly relaxed. Any response
    which vaguely resembles the standard behavior of the community is
    reinforced. When these begin to appear more frequently, a closer
    approximation is insisted upon. In this manner, very complex verbal forms
    may be reached. (pp. 29–30)

Another major topic of research was meaning.  behavioristic accounts of
meaning often emphasized associations among words.

Noble and McNeely (1957) constructed an index of the ‘‘meaningfulness’’ of
individual words by measuring the number of associations a person could
produce in a given time.
High-meaningfulness words such as kitchen were more easily learned ...  than
low-meaningfulness words such as icon (Underwood, 1966).

Osgood etal developed the semantic differential, a tool for measuring the
associative meanings of words by asking people to rate words on dimensions
such as good/bad and strong/weak (Osgood, Suci, & Tanenbaum, 1957).

divergence between psychology and linguistics


little interdisciplinary interest or activity for a period of several decades...

Striking example: linguist Leonard Bloomfield - once a student of Wundt’s and
published a book in 1914 with many Wundtian themes.

However, his linguistics text of 1933 took a more behaviorist view. In his
preface to the later book, Bloomfield tried to distance himself not only from
Wundt but from psychology as a whole:

    In 1914 I based this phase of the exposition on the psychologic system of
    Wilhelm Wundt, which was then widely accepted. Since that time there has
    been much upheaval in psychology; we have learned, at any rate, what one
    of our masters suspected thirty years ago, namely that we can pursue the
    study of language without reference to any one psychological doctrine,
    and that to do so safeguards our results and makes them more significant
    to workers in related fields. (Bloomfield, 1933, p. vii)

1950s - several meetings w psychologists and linguists - term
psycholinguistics coined (Osgood & Sebeok, 1965).  Many opposed the term,
e.g. Roger Brown: a ‘‘psycholinguist’’ sounded more like a deranged polyglot
than a psychologist interested in language (Brown, 1958),

Chomskyan linguistics


The second period of interdisciplinary psycholinguistics started
in the late 1950s, beginning with the emergence of the linguist Noam Chomsky.
Chomsky is generally regarded as the most influential figure in 20th-century
linguistics, and Newmeyer (1986) has characterized the Chomskyan influence
within linguistics as a revolution.

Chomsky's arguments against behaviorists view (Chomsky, 1957, 1959).

One behaviorist theory is the
associative chain theory: sentence consists of a
    chain of associations between individual words in a sentence.

Put another way, each word in a sentence serves as a stimulus for the next
word, and thus the entire sentence is produced left to right (at least for
European languages). Lashley (1951) had earlier argued against such a view,
claiming that there is something more to the structure of a sentence than the
associations between adjacent words.

Chomsky (1957) advanced this notion further. Consider the following
sentences:
	(5) Colorless green ideas sleep furiously.
	(6) Furiously sleep ideas green colorless.
	(7) George picked up the baby.
	(8) George picked the baby up.

Chomsky suggests that associations between words could not possibly explain
the existence of sentences such as (5).  associations between these words are
almost nonexistent, yet the sentence is syntactically cceptable.
But, if the words are presented backward, as in sentence (6), it is not a
sentence at all.

In (7) and (8), the sentences are synonymous, and while there is a relationship
between pick and up in these sentences, this is more complex
in (8) than in (7), because the words are separated.
how do we know in 8 that these words are part of a linguistic
unit, or constituent?
inguists call separate units, like those in sentence
(8), discontinuous constituents, and their existence suggests that there
are long-range dependencies among words in a sentence.

--> a theory that stresses a simple association between adjacent words is
    inadequate.

Poverty of Stimulus


Chomsky's primary argument is called the poverty of stimulus argument
(Chomsky, 1980). This argument states that there is not enough information in
the language samples given to children to fully account for the richness and
complexity of children’s language. Sentences (9) through (12) (from Caplan &
Chomsky, 1980) illustrate the point:

he/him can refer to John:
	(9) John believes he is incompetent.
	(12) John wants Bill to see him.
he/him cannot refer to John:
	(10) John believes him to be incompetent.
	(11) John wants him to win.

It is doubtful that anyone’s parents systematically distinguished between the
him in sentences (10) and (11) versus the him in sentence (12). In fact, most
people would not know how to explain such a difference. Still, we recognize
the difference and, moreover, can make a great number of other linguistic
discriminations about much more complex aspects of language that we are
similarly unable to explain in an explicit manner.  Chomsky’s argument is
this: The language children acquire is intricate and subtle, and the sample
of speech given to them during the course of language development is anything
but. Therefore, although parents may assist the child’s language development
in some ways and influence the rate of development somewhat, the pattern of
development is based not on parental speech but on innate language knowledge.

As Slobin (1971) puts it, a person who has learned a language
has formed something that is "psychologically equivalent" (p. 3) to a grammar.
Thus, psychologists became very interested in linguistics in general and in
Chomsky’s transformational grammar in particular (see Chapter 2).
     Slobin, D. I. (1971). Psycholinguistics. Glenview, IL: Scott,
     Foresman.

The psychologist George Miller created an important bridge between psychology
and linguistics by introducing psychologists to Chomsky’s ideas and their
psychological implications. Miller collaborated with Chomsky on several
articles and papers in the early 1960s (for example, Miller & Chomsky, 1963)
and was at the forefront of research during this period to determine the
psychological reality of linguistic rules (see, for instance, Miller & Isard,
1963).

Language development became popular... longitudinal investigations of child
language from the early 1960s (Braine, 1963; Miller & Ervin,
1964).  various ‘‘grammars’’ for child language were written, modeled
after adult grammars but differing in the specific rules (Bloom, 1970; Brown,
1973a). The major questions for language acquisition researchers were posed
in the following way: What set of rules governs the child’s developing
grammar, and when does this set develop?

Theoretical emphasis on role of innate factors. Eric Lenneberg was
influential in this - 1967 Biological Foundations of Language pulled together
evidence from aphasia, studies of delayed language development (for example,
mental retardation), and the available neurophysiological information into an
elegant argument for the role of innate factors in language development.

Another strong advocate of innate factors was David McNeill (1966, 1970), who
proposed a theory of development based on the concept of language universals.

Waning of Chomsky


...there are indications that psychological interest in linguistic theory has
waned. Reber (1987) examined the number of references to Chomsky in
psycholinguistic studies and found that they rose sharply in the late 1960s,
peaked in the mid-1970s, and then fell off by the early 1980s.
may reflect the trend among psychologists to shy away from directly
incorporating linguistic concepts into psychological research.

Reber cites several reasons:
* throughout the 1960s and 1970s linguistic theories underwent rapid and (to
  psychologists, at least) confusing changes (see Newmeyer, 1986).
  -> was difficult for psychologists to base their studies on any particular
  linguistic view, and some psychologists became wary of linguistics,
  preferring instead to develop a psychological view of language that was not
  tied to any specific linguistic theory.

  As Blumenthal (1987) has observed, there is a historical symmetry in these
  reactions — 70 years ago, linguists such as Bloomfield pulled away from
  psychology for much the same reasons.

* growing realization that the two fields were quite distinct in their
  methodologies, arising from the intellectual traditions of
  rationalism and empiricism - reminiscent of the dichotomy
  between heredity and environment, or nature and nurture: Rationalists
  emphasize the role of innate factors in human behavior, whereas empiricists
  stress the role of experience in behavior.
  MODES OF INQUIRY are also different. Rationalists emphasize the use of
  argument, whereas empiricists favor the collection of data as a means for
  evaluating hypotheses. For the most part, linguists approach language in a
  rationalistic manner; psychologists, even those who are sympathetic with
  the notion of innate factors, favor the empirical method.
		 (Pylyshyn, 1972, 1973; Watt, 1970).

  Watt, W. C. (1970). On two hypotheses concerning psycholinguistics. In
  J. R. Hayes (Ed.), Cognition and the development of language
  (pp. 137–220). New York: Wiley.

psycholinguistics today


although early psycholinguistics primarily focused on syntax, more recently
there has been an upsurge in interest in phonology, semantics, and
pragmatics. These developments have led to a more well-rounded field, with
research that cuts across these different areas (for example, Eberhard,
Cutting, & Bock, 2005).

more production than comprehension.  Are comprehension and production mirror
images of one another?  this view is misleading, as there are processes in
production that are not merely the reverse of comprehension (see chapter 8;
ref: Griffin & Ferreira, in press).

Also, neurolinguistic knowledge - effect of brain imaging - EEG, FMRI etc.

More integrative view.  Applying psycholinguistic research
to topics such as reading ( Just & Carpenter, 1987) integrates linguistic
theories of sentence structure, computer simulations of reading, and psychological
experimentation on eye movements.
Also applications to bilingualism (Bialystok, 2001),
language disorders (Tartter, 1998),etc.

2 Linguistic Principles


‘‘I don’t want to talk grammar. I want to talk like a lady in a flower-shop.’’
	—Eliza doolittle/Pygmalion (Bernard shaw, 1913/2000, p. 32)

Japanese, the basic word order is subject-object-verb (SOV). A simple
Japanese sentence (3) translates literally to
where hon means book and yatta means gave:
	(3) Taroo ga Hanako ni sono hon   o yatta. (Shibatani, 1987)
	    Taro  to Hanako    that book  gave,

Russian uses a rich set of affixes and prefixes [to indicate case roles], and
is much more flexible about word order. Viktor kisses Lena can be said in
many ways ([4]–[9], Comrie, 1987).

	(4) Viktor celuet Lenu.
	(5) Viktor Lenu celuet.
	(6) Lenu Viktor celuet.
	(7) Lenu celuet Viktor.
	(8) celuet Viktor Lenu.
	(9) celuet Lenu Viktor.

Turkish is a [richly agglutinated] language:
E.g. gel = come,
     gelemedim = I couldn’t come,
     gelemeyeceklermis = ~  [It was mentioned that] Those people won’t be
		able to come

In English, The elephant ate the peanuts must be marked for tense - whether
the event occurred in the past etc..

In Mandarin Chinese, indicating when the event occurred is optional.
In Russian, the verb would need to include not only tense but also whether
   the peanut-eater was male or female.
In Turkish, speakers must specify whether the eating was witnessed or just
   hearsay (Boroditsky, 2003).

Despite wide variations of structure across languages, some similarities.

e.g. Greenberg (1966) has discovered that every language contains declarative
sentences that express subject, verb, and object.

Duality of Patterning


a relatively small number of meaningless elements (e.g. spoken sounds) are
combined to form a large number of meaningful elements, the words.  (Hockett,
1966).
this form of duality does not appear to exist in animal communication.

Phones and Phonemes


Phones = speech sounds. Two sounds are different phones if they differ
    in a physically specifiable way. For example, consider the p in the
    words pill and spill.  There is a puff of air, known as aspiration,
    in pill that is not present in spill. You can tell the difference
    easily by placing a lighted match a few inches in front of your mouth
    as you pronounce the two sounds. Phones are indicated by brackets:
    The aspirated sound is symbolized as [p^h], the unaspirated as [p].

[TODO: candle test]

Phonemes are differences in sound that make a contribution to meaning;
    e.g. /b/ and /d/ are different phonemes: big vs dig.
    Phonemes may be thought of as categories of phones; each phone is a
    physically distinct version of the phoneme, but none of the differences
    between phones makes a difference to meaning.
    languages vary in their set of phonemes.
    In English, aspiration is not phonemic, although it
    is in Thai, which would represent the sounds as /ph/ and /p/.

Distinctive Feature:
The phoneme /b/ is similar to the phoneme /p/ except that the vocal cords
vibrate during the production of /b/ but not /p/. In distinctive feature
theory, contrasts are binary with the presence of the feature indicated by +
and its absence by -. The phoneme /b/ is said to be + voicing, whereas /p/ is
- voicing. In a similar vein, /b/ is + bilabial, which means that the sound
is articulated at the lips, and is + stop, meaning that the airflow from the
lungs is completely stopped during production. Distinctive feature theory (
Jakobson, Fant, & Halle, 1969) claims that these are independent units that
are combined to form phonemes.

Language Acquisition


Focuses on production.  Main points summarized at start:

* Children’s construction of language emerges from their understanding of
  communication prior to language. e.g. comprehension and production of
  gestures...

* children first acquire the sound system of their native language
  independently of meaning, and eventually merge it with communicative
  gestures to form productive speech.

* Development of one-word speech comprises two important developments:
  the acquisition of the lexicon and the use of single words to express
  larger chunks of meaning.

* Children’s first word combinations reveal a structure that is neither an
  imitation of adult speech nor fully grammatical by adult standards. With
  further development, children acquire the grammatical categories of adult
  speech.

* Early stages of acquisition are similar in signed and spoken languages.


Learning phonemic distinctions


Ability to distinguish phonemic distinctions from other languages declines in
strength during the first year of life.

Werker, Gilbert, Humphrey, and Tees (1981):

   compared 6- and 8-month-old infants from English-speaking communities,
   English-speaking adults, and Hindi adults. All three groups demonstrated the
   ability to distinguish between voiced and voiceless sounds, a distinction
   that is recognized in both Hindi and English. But only the Hindi adults and
   the infants were capable of distinguishing between pairs of Hindi sounds. It
   appears as if we lose some of our perceptual abilities over time.
   [from Traxler & Gernsbacher: Handbook of psycholinguistics
       While for English, three places of articulation are used for voiced stop
   consonants: labial, alveolar, and velar (e.g. /b/, /d/, and /g/,
   respectively), in Hindi four places are used: labial, dental, retroflex, and
   velar (e.g. /b/, /d/ (dental; dil), /D/, and /g/, respectively.) They created a
   synthetic series that varied perceptually from /b/ to /d/ (for native-English
   speaking adults) and from /b/ to /d/ to /D/ (for native-Hindi speaking
   adults).p. 179

 
  When hearing dental and retroflex Hindi stops [dil vs DAmar] 6- to
  8-month-old infants from English-speaking homes respond in a manner
  typical of native-Hindi adults and 11- to 12-mo Hindi infants Before
  they are a year old, English infants start treating both consonants
  the same.  p. 180 
]


amitabha mukerjee (mukerjee [at-symbol] gmail) 2012 Apr 20