Links to pages: 513, 514, 518, 520, 522, 523, 526, 528, 532, 535
Comments Welcome
George W. Grace
University of Hawaii
10 March 1989
(Revised 17 March 1989)
"[I]n knowing how to use their language, speakers know how to create and recognize associations between semantically interpreted sentences and particular types of situations." [Fillmore et al 1988: 502].
There are several kinds of indications that knowledge of one's language includes knowledge of associations between linguistic expressions (including entire sentences) and what we may call "situations". And some of these seem to indicate further that these associations are not round-about ones, but rather involve a direct connection in the speaker's mind between situation and linguistic expression. That is, there are observations which indicate the absence of--or at least exhibit no evidence of the presence of--what we may call "linguistic decoding" in the understanding of speech acts or of "linguistic encoding" in their designing. By "linguistic decoding" I mean a process in which one (1) determines the syntactic structure of the linguistic expression which one is undertaking to understand, and (2) on the basis of the meaning of the constituent lexical items and the "meaning" of that structure, determines the meaning of the expression as a whole. By "linguistic encoding", I mean something like the same process in reverse.
One indication of such a direct connection might be the fact that the syntactic analysis of a particular sentence is not always immediately accessible to speakers even though they may be able both to understand it and to use it appropriately. Even a linguist who is a native speaker of the language may, on occasion, have difficulty in deciding on one or another detail of the syntactic analysis of such a sentence, or different linguists who are both native speakers may find themselves in disagreement on details. These facts suggest that the kind of decoding process that I described in the preceding paragraph is not essential to understanding a linguistic expression--that there is another route from a linguistic expression to (in the words of Joseph Greenberg 1959: 74) "the situation to which it refers" which does not pass through the decoding step.
Another such indication is the ability of simultaneous interpreters to interpret while giving their attention to something else. And the something else may be verbal in nature--may require verbal activity on their part. There are numerous accounts of their writing letters or working crossword puzzles while interpreting. Nancy Schweda-Nicholson (personal communication) has even reported an incident in which the simultaneous verbal activity was itself oral. In the latter case the interpreter, while interpreting, was carrying on a conversation with a student. That is, he did not interrupt his interpreting activities, which is to say that he continued to take note of (if that is an accurate description), and to utter interpretations of, the remarks of the speaker being interpreted. But at the same time he also took note of what the student was saying, and replied to him whenever his interpreting duties did not require him to speak.
Thus, there is much anecdotal evidence to indicate that such an ability to convert source-language input to target-language output while one's conscious attention is engaged elsewhere is not at all uncommon among simultaneous interpreters. Although we cannot know what cognitive operations are being executed below the level of conscious awareness, it is difficult to imagine that they are of the complexity of the postulated decoding and encoding operations.
A third such indication is to be found in the reported response of people when they are asked to judge whether or not a string of words in their language constitutes a grammatical sentence--or at least their response in the more difficult cases.(1) My experience (and everyone I have talked to about it reports the same experience) was that I tried to think of a situation in which it would seem appropriate to use the string in question.
That experience seemed quite remarkable. It seemed remarkable first because it was so incongruous with the claim that we make such decisions on the basis of an internalized grammar. But it also seemed remarkable because it suggested some kind of direct linkage in our "minds" between sentence-sized linguistic expressions and "situations".
The fourth and final such indication I will mention here is the one which is most responsible for arousing my interest in the matter. It is the fact that proposed translations in which the target-language expression quite exactly matches the source-language expression in meaning (as determined by the meanings of the constituent lexical items and the grammatical constructions) are generally regarded as not being acceptable if the target-language expression is judged not to be what the target-language speakers would say in the same situation.
In Grace 1987b, I briefly discussed the dilemma posed by this fact--or, more precisely, by the following conjunction of facts (as I suppose them to be):
(1) The meaning of a linguistic expression such as a sentence is ordinarily assumed to be directly computable from the meanings of the constituent lexical items and the meaning(s) of the construction(s) in which they appear.
(2) It is commonly assumed that translation is concerned with likeness of meaning--that to translate something is to say something in the target language which has the same meaning. The natural interpretation of this specification is to produce a target-language linguistic expression which has the same meaning as the source-language linguistic expression.
(3) However, I have tried to show in Grace 1987b that where, given a source-language speech act to be translated, there is a target-language linguistic expression which is equivalent in meaning to the source-language linguistic expression, but that expression is judged not to be what target-language speakers would say "in the same situation", it is what they would say in the same situation that takes precedence.
Again, the fact that one is able to elicit judgments about what target-language speakers would say "in the same situation" seems in itself to be of particular interest. This means that the person making the judgment somehow determines, to his/her own satisfaction at least, how the relevant situation is to be defined. In fact, it seems to suggest some kind inventory of situations, or a theory which generates such an inventory.
It also means that for each of the situations in this inventory (or at least for very many of them) he/she feels capable of asserting with some confidence: (1) that other speakers of his/her language could be expected to speak in the situation in question, and (2) (within some limits) what linguistic expression they could be expected to use. This seems, therefore, to suggest something like an inventory of linguistic expressions (i.e., sentences or something of the sort) which could be mapped onto the inventory of situations. That is, it suggests that the linguistic expressions in question are known holistically (i.e., as lexical items are known) rather than, or in addition to, analytically.(2)
All of this seems, therefore, to suggest something like a dictionary which contains an inventory of linguistic expressions and an inventory of "situations" and which matches the members of the first inventory and the members of the second inventory in some approximation to a one-one mapping. But of course any such dictionary is very unlike anything which is supposed to be found in the linguistic competence of speakers of a language.
The statement by Fillmore et al which was quoted at the beginning of this Note also takes notice of our ability to recognize associations between situations and linguistic expressions. However, it appears as part of a description of "the standard idealization of the workings of a grammar", and its wording (I have in mind particularly the reference to "semantically interpreted sentences") is therefore designed to make it compatible with the syntax-semantics-pragmatics model of language. I am convinced that, on the contrary, our best hope of coming to understand something about the relation between situations and linguistic expressions lies in not being bound by any prior assumptions about the nature of language, and in particular in being willing to look at language without being bound by the syntax-semantics-pragmatics model.(3)
First of all, by the way, we should be clear that the linguistic expressions that we are talking about are sentence-level expressions, i.e., sentences or other longer or shorter expressions which are destined to serve as the vehicles of speech acts. Now if the association with which we are concerned between situations and linguistic expressions is a direct one, i.e., if we are to be obliged to deal with some sort of mapping between situations and such holistically-known linguistic expressions, we will presumably need a different understanding both of situations and of linguistic expressions.
I did discuss the problem of situations at some length in Grace 1987b, although of course that discussion is far from having completely disposed of the problem. There was, however, one point which needs to be recalled here. That point is that (cf. Grace 1987b: 460ff) if we are to have a theory of situations such that the use of particular linguistic expressions can be associated with particular situations, the situations in question will have to be subjective ones--that is, they will have in the final analysis to be states of the individual who is to speak. But we saw there that many subjective situations might be defined in terms of some external ("objective") situation plus a purpose or purposes of the individual with respect to that objective situation (to give a trivial example, the purpose of reporting the external situation in the most straightforward way).
If we are to consider the possibility that a language might be thought of as a kind of elaborated dictionary, it will be interesting to take a fresh look at the vocalizations of other primates. It has been said that these vocalizations differ from human language in that the former have "affective" content whereas the content of the latter is "referential". To my mind neither characterization has much validity. Although it is no doubt true that other species have nothing approximating the highly intellectualized uses of language--cf. my references below to "school language"--which have been developed in some recent human cultures (and which are particularly characteristic of what in Grace 1988 I call the "monoculture" of the modern world), it is very misleading to think of such intellectualized language as representing what is most characteristic in human language. (To suppose that it is is very reminiscent of the old assumption that Western intellectuals were somehow normal, and that children, the mentally deranged, and "primitives" all showed the same kinds of intellectual deficiencies--when what was perceived as deficiency was nothing but essential humanness).
Anyway, I am aware of no obstacle, on present evidence, to describing the vocalizations of other primates, such as East African vervet monkeys, as responses to subjective situations. This description, I believe, could apply equally well to the "alarm calls" (which, some argue, qualify as "referential") and to the "grunts" (cf. Seyfarth 1984 and references cited there).
It is actually very difficult to be sure just how communication systems of other species work (and of course I am also convinced that we currently operate with quite distorted ideas of how human language works as well). However, suppose the suggestions being considered here are valid: that both primate vocalizations and human language are, in their essence, mappings between (subjective) situations and vocal expressions. The hypothesis that this is so would put the problem of the origin and evolution of language suddenly in a very different light. We would no longer have to account for an almost mystical metamorphosis, but might instead find ourselves attempting to specify the details of a gradual progression in complexity.(4) Although "creation scientists" might find less comfort in such a picture of the development of language than in the present one, for most of us surely the prospect would be quite appealing.
The proposal is that these vocalizations and typical (sentence-level) linguistic expressions of human language correspond to, and signal, subjective situations. But I also suggested that subjective situations can probably, at least in a great proportion of cases, be defined in terms of an objective situation plus a purpose or purposes of the individual with respect to that objective situation. Evolution within the animal kingdom has led from organisms capable of recognizing only a very few differences of environment (cf. the discussion of the Umwelt of the tick in Uexküll 1957)--that is, possessing only a very limited inventory of possible external situations--to others which are capable of making a very large number of subtle discriminations (cf. the discussion of the increase in capacity for integrating sensory information in Jerison 1973). (It is not impossible, then, that the differentiation of the perceived external environment has proceeded more rapidly than the differentiation of internally-generated states--in other words, that human discourse has actually come to be more oriented to the external world than that of other species. However, human discourse has also led to the ability to construct fictive external situations.)
If the scenario suggested here is valid, holistically-known expressions are phylogenetically prior to analytically-known expressions. They are ontogenetically prior as well. That has long been known, of course. But what is significant, as was made apparent particularly by the work of Ann Peters (Cf. Peters 1977, 1983) is the way in which the distinction between word-level and sentence-level signs comes about. It is not simply a matter of the child's beginning with words and then, at a later stage, acquiring the rudiments of syntax required to combine words into longer utterances. It can also proceed in the opposite direction, so that the original unit is analyzed as having been a gestalt (a holistically-known sentence-like unit), with its component parts (word-like units) the secondarily-arrived-at units.
But this view of language in which dictionary-type knowledge (even of sentence-level signs) is primary and grammar-type knowledge secondary is in diametric opposition to the standard view. If the proposed view has any validity at all, how can the standard view ever have been conceived? I think the answer lies partly in the difference between what we may call "school language" and the vernacular.
Frances Syder (Syder 1983) and Andrew Pawley (Pawley 1984, Pawley and Syder 1983b) have pointed out that the language (and use of language) taught in the schools is different from the vernacular language of ordinary life. (Their remarks are directed particularly to the schools of New Zealand, but are equally applicable to those of many other countries--presumably all countries in approximately the measure to which they have been affected by the monoculture). In fact, the difference is so great, according to some observers, that the vernacular language unfits the child for school learning, and/or school language unfits people for life after school (Pawley 1984: 25).
Where did school language come from? The Latin of the Middle Ages and Renaissance may have played a central role. Walter Ong (1977: 25) wrote of it, "Learned Latin [sc. the Latin of the period from around A.D. 550 to the present--GWG] has not been inherited from within the family and has normally been used exclusively to deal with tribal and public affairs rather than domestic affairs. That is, it has been used for more or less abstract, academic, philosophical, scientific subjects or for forensic or legal or administrative or liturgical matters."
Further (1977: 27-28), "This is a strange situation for a language. Latin was distanced--alienated--not from day-to-day life, for it was of the substance of daily life for lawyers, physicians, academic educators, and clergymen, but from the psychological and psycho-somatic roots of consciousness. It no longer in any sense belonged to mother, it did not come from where you came from."
And (1977: 36) "Modern science only gradually became viable in the vernacular atmosphere as it transformed this atmosphere by injecting it with Latin terms and forms of thought."
The (certainly speculative) picture that I would suggest, then, is that in the European civilization of the Renaissance school language existed in the form of Latin. It was taught as a second language (in fact, I would like to suggest that the Latin instruction of the Middle Ages and Renaissance and the materials prepared for use in that instruction have provided the models used by later generations both for second-language instruction and for linguistic description [the grammar-lexicon model]).
Anyway, to focus on England (although there was a similar course of events elsewhere in Western Europe at about the same time, and eventually everywhere the monoculture has reached), there was the school language--Latin--and the vernacular--English. Latin was "regulated"--it was understood to be governed by rules, rules which were discussed and formulated in writing. The vernacular was also written, but in something of a catch-as-catch-can fashion without fixed rules (e.g., note the inconsistencies in the spelling of that period). And, in particular, the uses--the roles--of school language and the vernacular were quite separate.
But eventually the functions of Latin began to be assumed by the vernaculars. This apparently came about primarily through translations of Latin works into vernacular languages after the invention of printing. The motive seems simply to have been to sell more books to more customers by making them available in a language which the potential customers knew. In the process English and other vernaculars were forced to accommodate translated Latin. Once they had been thus adapted to translate Latin works, they could then be made to serve for writing original works on Latin subjects (or what had thitherto been Latin subjects). And in so doing they were also obliged to mimic--to calque--Latin ways of talking (cf. Grace 1987a: esp. 92-107) about these subjects.
In the course of these changes, rules were laid down for them as well--their spelling was standardized, grammatical rules were prescribed, and dictionaries were made. What in fact happened was that languages like English came to encompass on one linguistic chassis, continuations of both the school language and the vernacular.
But as the school language assumed its place as a part of English, it brought its prestige with it. It is not surprising that the prestige should go to school language. There had always been an association between the school language and education. School language was, and remains, a badge of the status associated with education. But a corollary of this prestige was that where a choice had to be made, "good" English was school English. In fact, school English came to be regarded as the real English--any deviation from it was a deformation and not representative of the English language.
Thus, a skilled command of school language connotes education and carries prestige. Conversely, the specialists in language within the society even today are among its most highly educated members. They tend to be exceptionally skilled in the use of the school language (and no doubt to take justifiable pride in that fact).
I am proposing that for all of these reasons the school language assumes special prominence in the lives of those whose profession it is to study language, and that this has tended to give some characteristics of school language particular prominence in the standard view of the nature of language and of the nature of particular languages such as English. One such characteristic is the concern with grammatical rules as the primary factor governing correct knowledge and use of the language. A second is the ideal for language production that each sentence should be a carefully-designed one in which every word has been independently chosen. Both of these, of course, represent the standard view of language and disagree with the hypothesis under contemplation here.
The hypothesis that I want to propose for your consideration, then, is that language began as a set of holistically-known (in fact, presumably unanalyzable) linguistic expressions. It was the function of each of these expressions to constitute an utterance (to serve as the vehicle of a speech act), and the "meaning" of each was (to call attention to--to signal--to "signify") a subjective situation. And this, so the hypothesis goes on, is still the essential nature of language today. However, the system of situations that require to be signalled has grown to be extremely extensive and complex. And, accordingly, the system of linguistic expressions has likewise become very complex. Sentence-level linguistic expressions are now typically analyzable, in fact, their composition is often quite complex.
School language constitutes a sort of ideal, and an ideal for composition in school language is that each sentence is constructed from scratch. Each word is individually selected, and placed within a grammatical construction which is so designed that the whole specifies exactly the intended "meaning" (in the words of Joseph Greenberg 1959: 74, "analytically specifies the situation to which it refers"). What I am proposing is that ordinary language, vernacular language, comes as near the exact opposite of this as it can--as near its opposite as the complexity of the human condition permits.
But it cannot come too near because the array of situations to be distinguished is too great. The hypothesis proposed here is that the evolution of language was in fact the evolution of ways of providing new signals for new situations. It was, in fact, a response to a rapid increase in the capacity of pre-humans and humans for processing sensory information. This increased capacity greatly increased the human ability to identify and respond appropriately to subtle environmental differences--in other words, to distinguish more and more different situations.
This development would presumably not of itself necessitate a concomitant enrichment of the system of vocalizations associated with (signalling) the situations. However, it seems obvious in retrospect that such an enrichment was advantageous. Once a new situation had been isolated, having a specific signal dedicated to it would presumably serve not only as a means of signalling it to others but also as a means of fixing it in one's own memory for future reference. However, our hypothesis does not require that we explain why language-as-we-now-know-it developed. It is a fact that it did. The hypothesis concerns what language-as-we-now-know-it is really like (which should be enough of a problem for one Note).
Now what happened, according to the hypothesis, is that new vocalizations (or let us henceforward think and speak of them as sentence-level "linguistic expressions") were invented on the basis of old ones.
One point which seems clear is that what I have been calling an "inventory" of situations is not simply an unstructured list. For one thing, it contains many groupings of similar situations which are signalled by correspondingly similar linguistic expressions. An obvious example is a grouping where the differences are signalled by different forms in a paradigm such as a verb conjugation. E.g., Latin amo, amas, amat, etc. are all capable of functioning as sentence-level expressions signalling situations, but the situations they signal all belong to a single grouping the members of which are systematically related to one another. Many such groupings, associated with paradigmatic sets of linguistic forms, could be identified (presumably in any language if the notion of paradigm is suitably construed). Such groupings would not be restricted to one-word sentences; I see no reason not to recognize such English sentences as "I love you.", "He loves her.", and even "John used to love Mary." as all belonging to a more general kind of situation-grouping. In fact, from the point of view of the hypothesis which we are considering, grammar presumably was developed to fill the function of signalling new situations as they arose. And what more efficient system of signalling them could be developed than one which represented each new one in terms of existing ones to which it was similar?
But groupings based on such linguistic paradigms are not the only ones that can be recognized. Here the work of Andrew Pawley is most helpful. He and Frances Syder have collected extensive recordings of natural conversation. They were impressed by the large number of "lexicalized sentence stems" (Pawley and Syder 1983a: 191) and "speech formulas" in general (e.g., Pawley 1985, 1988--cf. also Pawley 1986). I assume that all such formulas represent not single situations but groupings of similar situations.
In Pawley and Syder's definition (1983a: 210), "A sentence stem consists either of a complete sentence, or, more commonly, an expression which is something less than a complete sentence. In the latter case, the sentence structure is fully specified along with a nucleus of lexical and grammatical morphemes which normally include the verb and certain of its arguments; however, one or more structural elements is a class, represented by a category symbol such as TENSE, NP, or PRO." They state (1983a: 192) that the "stock of lexicalized sentence stems known to the ordinary mature speaker of English amounts to hundreds of thousands".
Pawley 1985 uses the term "speech formula", which I take to represent a more inclusive concept that includes sentence stems. He gives the following definition (1985: 88): "A speech formula is a conventional pairing of a particular formal construction with a particular conventional idea or idea class." I believe that his "idea or idea class" can--at least for the present--be equated roughly with my "situation". (I may note that in Pawley 1988, he describes a formulaic construction as pairing a grammatical formula with a "discourse function" [rather than an "idea"]).
He goes on to say (1985: 88), "Not only is the idea familiar, but it has a certain independence from form, perhaps being expressible by a number of different forms without losing its status as 'the same idea'." Although what I have been saying here may have seemed to suggest that the association of situations with linguistic expressions was a one-one matching, I do not want to imply that at all. I do not see any disagreement between what he proposes and the hypothesis put forward here. (His [Pawley 1988] concept, "semantic formula" is of interest in this connection, but I will not attempt to pursue it here.)
Where the formula is a construction, the degree of lexicalization may vary. He distinguishes three degrees of lexicalization or lack thereof. I will quote further: "A (formal) construction consists of two or more syntactic units standing in construction, e.g. as a complex word, phrase, clause, etc. In a formulaic construction, the syntactic units may be (i) lexically specified in all cases, i.e. the construction is made up of a fixed sequence of morphemes, such as A stitch in time saves nine, or Every man has his price, (ii) lexically specified in some cases, as in the formulas Long live NP! and The ADJ-er (X), The ADJ-er (Y), discussed below, or (iii) lexically unspecified, i.e. made up of syntactic categories such as V, N, NP, etc. In the last case, however, one or more of the syntactic categories is lexically very restricted; it represents a small class of lexemes, usually one that forms a semantic set, as in the time-telling formula M PREP H, discussed below." (1985: 88-89).
"Grammatical constructions" in George Lakoff's (1987) conception seem to be speech formulas (formulaic constructions) in Pawley's sense. Grammatical constructions are described (Lakoff 1987: 464) as "direct pairings of parameters of form with parameters of meaning." Lakoff's treatment is of particular interest because it suggests how new formulas may be created.
Some constructions are more central than others. The peripheral constructions are based on the central ones to which they are linked. Grammatical constructions are not predictable from general syntactic, semantic, or pragmatic principles, but they are motivated by such principles.
This conception of grammatical construction seems quite valid to me. I propose, then, that one solution to the problem of designing a new linguistic expression to fit a newly isolated situation has been to create a new grammatical construction based on an existing construction which was appropriate to similar situations, and such that the difference between the old construction and the new one is motivated by the difference between the original situations and the new one.
Much of the hypothesis that I am tentatively putting forward here is sketchy, and it is difficult to specify just which parts of it can be said to be supported by existing evidence. However, I am convinced that there is an association between something that I am here calling "situations" and linguistic expressions which has not been given sufficient attention in existing theories of language.
I'm not sure to what extent this hypothesis may be compatible with the syntax-semantics-pragmatics model of language. What I would, of course (cf. endnote 3), like to see is for the syntax-semantics-pragmatics model to be subjected to an open-minded re-assessment.
I would suppose that in the hypothesis that a language is essentially a mapping between situations and linguistic expressions, grammatical regularities would be explained as the by-product of the processes of creation of new linguistic expressions to signal new situations. It is not clear to what extent these regularities might be expected to combine to form overall syntactic systems. That is, it is not clear what expectation there should be of the program of writing a single set of syntactic rules for a language ever being completed. I would guess the prospects would be better for languages that incorporate something like the "school language" described here than for others.
Wittgenstein's metaphor seems to me to be very apt here. He wrote (1958: 8), "Our language can be seen as an ancient city: a maze of little streets and squares, of old and new houses, and of houses with additions from various periods; and this surrounded by a multitude of new boroughs with straight regular streets and uniform houses."
(I take the new boroughs to be the contribution of "school language".)
1. Of course, of all of the strings which could be assembled out of the words of a particular language, say English, the vast majority would be obvious monstrosities patently at odds with the structure of the language at almost every point. On the other hand, there would be some which were recognizably familiar. Some of these would be sentences which one actually remembered from having heard them used repeatedly. In other cases, I'm not sure exactly what is involved--maybe one has actually heard them, although one can't exactly pinpoint the occasion, or maybe they are just so similar to remembered sentences or are appropriate for such familiar situations that they might be described as in some sense "expectable". (For example, what is the status of the expressions cited in Reddy 1979 [passim, and especially 311-24] or Lakoff and Johnson 1980 [passim]? I presume that we haven't actually heard all of them before, but if not, what is the basis of their recognizability?)
Anyway, as research within the TG tradition developed, more and more strings were produced which might be described as "marginal". When we were asked whether or not one of these strings was "grammatical" or was at least something which "could be said", often the answer was not immediately apparent. They appeared to be neither clearly ungrammatical nor clearly all right. We had to search for an answer.(Back up)
2. In Grace 1981 (esp. 19-20) I proposed the terms "analytic" and "holistic" for two modes of knowing (the aspects of) one's language. The analytic mode is "the way of knowing structural principles, generalizations". It is "especially prominent in the speaker's knowledge of those aspects of a language that are generally described in grammars" (1981: 20). The holistic mode is "the way of knowing a linguistic form immediately and as a unit"--i.e., the mode of knowing typical of lexical items. I proposed there "that a speaker potentially may recall (i.e., know holistically) any collocation, sentence, etc. which he has ever heard." (Back up)
I do think that this set of assumptions about the nature of language is a cause for concern. Or rather it is not the assumptions themselves that are the cause for concern, the concern is with the fact that they are put forward as representing the true nature of language. No matter how useful these assumptions about the nature of language (and the way of looking at and talking about language which they lead to) may be for investigating the questions which are regarded as most central to linguistics, they can be quite harmful if they are imposed on attempts to answer questions for which they are not suitable. I have at one time or another believed that they were a serious impediment in studies of such diverse matters as animal communication, the evolution of language, aphasia, translation, language "contact" (indeed, of linguistic change generally), language learning (both first and second language), and in fact of all fields where what is known (or assumed to be known) about language is applied to the understanding of other phenomena in which language is concerned.
However, this is not the time or place to argue the harmfulness, or even just the inherent implausibility, of the syntax-semantic-pragmatics model. I am only concerned here with being able to discuss the relations between situations and linguistic expressions without having to assume beforehand that these relations must be consigned to some peripheral "pragmatic" component of the language.(Back up)
4. But of course details of the continuity proposed here are extremely sketchy. I am not sure how this continuity proposal might be reconciled with the bioprogram proposal of Bickerton 1981, which contains important grammatical elements, but any serious attempt to develop the continuity proposal further would have to confront this problem. (Back up)
Bickerton, Derek. 1981. Roots of language. Ann Arbor: Karoma Publishers, Inc. (Back up)
Fillmore, Charles J., Paul Kay, and Mary Catherine O'Connor. 1988. Regularity and idiomaticity in grammatical constructions: The case of let alone. Lg. 64: 501-38.(Back up)
Grace, George W. 1981. An essay on language. Columbia SC: Hornbeam Press. (Back up)
Grace, George W. 1987a. The linguistic construction of reality. London: Croom Helm. (Back up)
Grace, George W. 1987b. "What they would say in the same situation". Ethnolinguistic Notes, Series 3, Number 31. Printout. 19 pp.(Back up). Also internet World Wide Web page (Click Here).
Grace, George W. 1988. The idea of a theory of translation: On shared and unshared cultural backgrounds. Ethnolinguistic Notes, Series 3, Number 33. Printout. 10 pp.(Back up). Also internet World Wide Web page (Click Here).
Greenberg, Joseph H. 1959. Language and evolution. In Betty J. Meggers (ed.). Evolution and anthropology: A centennial appraisal. Washington DC: Anthropological Society of Washington. (Back up)
Jerison, Harry J. 1973. Evolution of the brain and intelligence. New York: Academic Press.(Back up)
Lakoff, George. 1987. Women, fire and dangerous things: What categories reveal about the mind. Chicago: University of Chicago Press. (Back up)
Lakoff, George and Mark Johnson. 1980. Metaphors we live by. Chicago: University of Chicago Press.(Back up)
Ong, Walter J. 1977. Interfaces of the word: studies in the evolution of consciousness and culture. Ithaca NY: Cornell University Press.(Back up)
Pawley, Andrew. 1984. School English is nobody's mother tongue: Reflections on vernacular and school-acquired language. In Allan Berry (ed.). Communication: Papers from the twentieth extension course lectures held in the Auckland Institute and Museum, April 3-4, 1984, pp. 25-41.(Back up)
Pawley, Andrew. 1985. On speech formulas and linguistic competence. Lenguas Modernas 12: 84-104.(Back up)
Pawley, Andrew. 1986. Lexicalization. In: Georgetown University Round Table on Languages and Linguistics 1985. Edited by Deborah Tannen and James E. Alatis. Washington DC: Georgetown University Press. 98-120.(Back up)
Pawley, Andrew. 1988. Formulaic speech. (Draft of article for the Oxford International Encyclopaedia of Linguistics). ts.(Back up)
Pawley, Andrew, and Frances Hodgetts Syder. 1983a. Two puzzles for linguistic theory: Nativelike selection and nativelike fluency. In Jack C. Richards and Richard W. Schmidt (eds.). Language and Communication. London and New York: Longman, pp. 191-225.(Back up)
Pawley, Andrew, and Frances Hodgetts Syder. 1983b. Natural selection in syntax: Notes on adaptive variation and change in vernacular and literary grammar. Journal of Pragmatics 7: 551-79.(Back up)
Peters, Ann M. 1977. Language learning strategies: Does the whole equal the sum of the parts? Lg. 53: 560-73.(Back up)
Peters, Ann M. 1983. The units of language acquisition. Cambridge, etc.: Cambridge University Press.(Back up)
Reddy, Michael J. 1979. The conduit metaphor--a case of frame conflict in our language about language. In Andrew Ortony (ed.). Metaphor and thought. Cambridge: Cambridge University Press, pp. 284-324.(Back up)
Seyfarth, Robert M. 1984. What the vocalizations of monkeys mean to humans and what they mean to the monkeys themselves. In Rom Harré and Vernon Reynolds (eds.). The meaning of primate signals. Cambridge: Cambridge University Press, pp. 43-57.(Back up)
Syder, Frances Hodgetts. 1983. The fourth R: Spoken language, English teaching and social competence. Palmerston North, New Zealand: Frances Hodgetts Syder.(Back up)
Uexküll, Jakob von. 1957. A stroll through the worlds of animals and men: A picture book of invisible worlds. (translated and edited by Claire H. Schiller--original German edition, 1934). In Claire H. Schiller (ed.). Instinctive behavior: The development of a modern concept. New York: International Universities Press, pp. 5-80.(Back up)
Wittgenstein, Ludwig. 1958. Philosophical investigations. Translated by G. E. M. Anscombe. 2nd Ed. Oxford: Basil Blackwell.(Back up)
| Home Page | The Ethnolinguistic Notes | The Ethnolinguistic Notes, Series 1 and 2 | Ethnolinguistic Notes, Series 3 | The Ethnolinguistic Notes, Series 4 | Reflections: Language Evolution |
| Reflections: Knowledge of Language | Personal Page | The Human Predicament | Why Write Unpublishable Things? | Modest Proposals | Odds and Ends | Pictures |
First put on the Web on 18 February 1997