The following was
published in the journal Historiographia Linguistica 29: 245-261
SEARCHING FOR MEANING IN
In the fall of 2004, counting from the year I entered graduate school, I will have participated in the field of linguistics for half a century. Perhaps that earns me the right to reflect a little on my involvement in this field, but more importantly to say a few things about trying to understand the nature of language and how it relates to the totality of human experience, even making a few suggestions about lines of research that might be worth pursuing in the years to come.
Linguists occasionally talk about why they became linguists, and often they describe a childhood environment where several languages were in the air. They at least heard and may have spoken Yiddish, Norwegian, Japanese, or whatever. I was born in 1927 in Cambridge, Massachusetts, but my family moved from there when I was still a baby, and until I was ten we lived in a small Massachusetts town called North Attleboro, where diversity was a matter of being Methodist or Episcopalian. I had a classmate named Sarkis Meranian whose parents probably spoke Armenian, but I didn’t think about it at the time. Still, my father had studied Greek for eight years and Latin for five, first at Boston Latin School and then at Harvard, and I found his Greek dictionary so fascinating that one day I took it to school for what was later called Show and Tell. My teacher was so impressed that she had me take it around and show it in other classrooms. I was fascinated, too, by a French textbook my mother had used in high school. I read that the “L” at the end of table was “weakly pronounced,” and I wondered what that could mean. Maybe there was some interest in language or languages hidden in my genes, even though at first it didn’t get much encouragement.
When I was ten we moved to Lynn, an industrial city north of Boston, where the population was as diverse as it had been homogeneous in North Attleboro. The parents of my friends spoke Italian, Yiddish, Greek, or Polish, and linguistic diversity was hard to miss. In school I took pleasure in learning Latin and a bit of French, but in those years I was more caught up in playing the piano, clarinet, and saxophone. After high school and a stint in the Navy during and immediately after World War II, I entered Yale, where I majored successively in music, architecture, and German literature. At that time Yale offered only one undergraduate course in linguistics, taught by Bernard Bloch. I didn’t take it but a friend, Robbins Burling, audited it, and it was through him that I learned there was such a field, the idea of which I found strangely appealing. When I graduated in 1950, I wanted to see more of the world and found myself for several years in Switzerland, where multilingualism was everyone’s birthright. Even before I went there I stopped off in New Haven to see what it would take to do graduate work in linguistics. I talked with Franklin Edgerton, who told me to forget it unless I had an independent income, and with Bloch, who told me to read Bloomfield’s Language. In the end I followed Bloch’s advice but not Edgerton’s, enjoying Language on the balcony of my chalet in Bern. In the fall of 1954 I returned to Yale as a graduate student.
The Yale linguistics program (it was not yet a department) had two strong orientations. One was Indo-European studies, which were thought to be the indispensable core of linguistic knowledge. The program laid out for me—I had little choice—included Sanskrit (with Paul Tedesco and Paul Thieme), Gothic and Runes (with Konstantin Reichardt), Hittite (with Albrecht Goetze), and Greek and Comparative Indo-European (with Ralph Ward). I remember suggesting to Bloch that it might be useful as background for my dissertation to take Floyd Lounsbury’s course in the Ethnology of the Eastern Woodlands. Bloch steered me into Hittite instead. There was a clear notion of what was important.
The other orientation at Yale was what later came to be called American structuralism, although at the time we just thought of it as the right way to do linguistics. Yale was the strongest of strongholds of Post-Bloomfieldian thinking, and as long as I was there I accepted it as the only proper mindset. I assimilated the view that Sapir was a genius but Bloomfield was right. Bloch was our theoretical anchor, and his influence extended well beyond Yale through his editing of the journal Language and his contacts with the big names in the field. Bloomfield had died while I was still an undergraduate, but his ghost was a constant presence. Language was the bible that guided our thoughts and deeds, and it never occurred to me then to see things differently.
I wasn’t drawn to Indo-European as a career, and because both Sapir and Bloomfield had worked with American Indian languages, I thought it might be exciting for me to do that too. It was thus that I met Lounsbury, whose subsequent influence on many aspects of my life was profound (Chafe 1999). He had extended his early contacts with the Oneida language in Wisconsin to greater or lesser amounts of fieldwork with the other extant Iroquoian languages, but less with Seneca than the others, and that language became the natural choice for me. He quickly introduced me to the dean of Iroquoian ethnology, William Fenton, then director of the New York State Museum in Albany, who, I was later intrigued to learn, had taken a course in phonetics with Sapir at Yale in the early 1930s.
It happened that Fenton was just then sending out specialists of various kinds to conduct salvage surveys in and around the Seneca reservation on the Allegany River in western New York. For many years the U.S. Army Corps of Engineers had been salivating over the prospect of a dam that would be located just downstream from the reservation, flooding a significant part of Seneca land. Its construction seemed imminent, and ecological surveys of the area were a priority. Fenton had a special interest in the Seneca language, and may have thought that it too was threatened by the dam. In any case he signed me on as a temporary employee of the museum, and my work with Seneca speakers was enriched through contacts with botanists and zoologists. I worked for four summers, 1956-59, on all three Seneca Reservations—Allegany, Cattaraugus, and Tonawanda—and by 1958 I had completed a structuralist dissertation on Seneca morphology, later published together with a dictionary (Chafe 1967). I was privileged to work with and get to know some remarkable members of the Seneca community, among others Solon Jones, Corbett Sundown, and Edward Curry, whose oratorical skills will not be matched again. In those days Seneca was widely used, though one could already see it losing ground to English, especially among younger people. I thus became a witness to the disappearance of languages we now see occurring everywhere, with tragic effects for the communities concerned, a radical diminution of the basic subject matter of linguistics, and ultimately a grievous loss for all humanity.
Ever since my early fieldwork I have been convinced of the value of working in depth with one or more languages radically different from one’s own. Getting one’s hands dirty observing real speech, even in a familiar language but especially in an unfamiliar one, brings insights obtainable in no other way. I learned, too, that uncovering the mysteries of a language through fieldwork is the purest application of the scientific method. Observing a language in use, imagining explanations for why it has the shape it has, and seeing those provisional explanations confirmed, disconfirmed, or modified by further observations brings unequaled thrills of discovery, an experience that becomes less available each time a language dies.
In 1958 I needed a job. As I sat in Bloch’s office discussing what I might do, he picked up the phone and called George Trager, with the result that I spent a year in Buffalo teaching German and absorbing the atmosphere of the special brand of Post-Bloomfieldian thinking associated with Trager and Henry Lee Smith. Living close to the Senecas was a benefit, as was the chance to meet regularly with the ethnologist Elisabeth Tooker and the archeologist Marion White to discuss Iroquoian studies from our complementary perspectives. There was, however, no plan to work me into the Buffalo linguistics program (Trager and Smith wondered a little about me), and I was lucky the following year, with the support of William Sturtevant, to find employment in the Bureau of American Ethnology of the Smithsonian Institution. I was in a vague sense a successor to the legendary John Peabody Harrington, who had retired years earlier. I occupied his former office but we never met, and he died soon after in Santa Barbara.
Sapir had suggested that the Iroquoian languages might bear a special remote relationship to languages of the Caddoan family. Very little was known of the latter, and in the fall of 1959 I made an initial survey of the four Caddoan languages still spoken: Pawnee, Arikara, Wichita, and Caddo. I decided to concentrate on Caddo, in part because I found the language unusually appealing, in part because I met a gifted Caddo woman named Sadie Bedoka Weller. Between 1961 and 1965 I spent four summers in and around Anadarko, Oklahoma, working with her and other Caddo speakers. There was a change of pace in the summer of 1963, when I took part in a seminar on computational linguistics at the Rand Corporation in Santa Monica, designed to teach us what computers could do and how to use them. On the faculty were Charles Hockett and Sydney Lamb, and my fellow participants included various important linguists of the day. I discovered that I had a talent for computer programming, though it was a very different enterprise then than now.
The job at the Smithsonian was in some ways ideal, allowing almost unlimited time for research and fieldwork, and providing an unparalleled library and archive of Indian language materials. I missed, however, the stimulus of students, and I compensated to some degree by teaching part-time at both Georgetown and the Catholic University of America. I remember with special fondness two field methods courses, one with Winnebago and one with Cherokee. But when the chance came to move to Berkeley in the summer of 1962 the decision was not difficult. The Berkeley Linguistics Department was still basking in its golden age of Indian language studies under the guidance of Mary Haas. It seemed a natural place for me to be, in spite of the fact that my work was centered in New York and Oklahoma, not California. Before long I was fortunate to be joined in Caddoan studies by two Berkeley students, David Rood and Douglas Parks, the former tackling Wichita, the latter Pawnee and later Arikara.
During my last year at Yale I heard a talk by someone named Noam Chomsky on something he called transformational grammar, in which Bloch had taken an interest. Thus began my acquaintance with a way of thinking about language that left me forever working on the margins of the discipline. I think I know what it was like to be an atheist in medieval Europe, or a believer in the importance of consciousness and mental imagery in the heyday of behaviorist psychology. Chomsky’s vision of language struck me as curiously superficial, and I didn’t think it would win any wide acceptance, hardly expecting the accelerating pace with which it captured the field during the 1960s. Because everyone else was doing it I even tried it on for size myself, but soon decided that my initial instincts had been correct.
Sometime in the early 1960s it became clear to me that meanings were as important to the structure of language as sounds, and perhaps more important—that language was fundamentally a way of associating meanings with sounds, the meanings determining in the first instance the shape that language took. To make this association with sounds possible, and to promote communication, it was necessary for each language to have its own way of organizing meanings, just as each language has its own way of organizing sounds. I first tried to articulate that view as early as Chafe (1962), and ever since then I have been trying to expand on it.
There was one respect in which I was able to connect with the generative movement: I found validity in the notion of a deep structure underlying the surface structure of a language. I differed in seeing this distinction in the context of idioms, with their deep meanings expressed in surface forms. I tried to show how idioms proved the dependence of syntax on semantics (Chafe 1968). Much later, influenced by more recent studies of idioms, I recognized what I have come to call shadow meanings—the literal meanings of at least some idioms that remain lurking in the shadows, influencing to some degree the way people think when idioms are used. Later, too, I realized the importance of relating idiom formation to the grammaticization of inflectional meanings (Chafe in press a). But in the 1960s I focused on the notion that if deep structure were to make any sense, it would have to be equated with semantic structure, and that was the perspective I tried to develop in Chafe (1970a).
The timing was bad. My attempt to base grammar on semantics just happened to coincide with the rise of generative semantics, an offshoot of generative grammar, which I thought was right in associating deep structure with semantics but wrong in the way it conceived of semantics itself. To oversimplify the difference, whereas I thought that syntax should be more like semantics, the generative semanticists conceived of semantics as structured in a form still dictated by generative syntax. For a moment I thought I could join them, but I soon gave up when I heard about lexical decomposition and realized how beholden the generative semanticists still were to what I saw as a deficient view of language. While I was a visiting professor at Cornell in 1968 I gathered material for a sketch of another Iroquoian language, Onondaga, spoken not far from Ithaca, and at first I titled it A Generative Semantic Sketch of Onondaga. With my disillusionment I scratched out that title and replaced it with A Semantically Based Sketch of Onondaga (Chafe 1970b). Generative semantics failed to capture the high ground (Lakoff 1989, Huck and Goldsmith 1995), and of course my own solitary attempt had little influence, or so I thought. In recent years a number of people have surprised me by saying that my 1970 book made a difference to them. With hindsight I can fantasize that a different power structure in the field would have encouraged me to pursue that direction further. I have no doubt in any case, apart from my own involvement, that linguistics without generativism would have enjoyed a more productive history from the late 1950s until now, that today we would be able to boast of more substantive accomplishments.
At the beginning of the 1970s I mounted a project to investigate semantic prerequisites to machine translation. Although computers in their early stages had aroused an unrealistic enthusiasm for both artificial intelligence and machine translation, interest in the latter was waning, but in spite of that the U.S. Air Force was still willing to support related research to some degree. I was sure that high quality machine translation was in principle impossible, but I rationalized our effort with the thought that some kind of machine-aided translation might be worth pursuing. At the very least our project could examine what its underpinnings might be. We didn’t get very far before funding ran out, but it was an effort that led to the more satisfying Pear Stories project in the mid-1970s.
I had become interested in what I saw as an unconformity between thought, as it was organized by language, and total thought, which I saw as manifested not only in language but also in mental imagery and emotions. I had conducted some little data-gathering tasks in which I showed home-movie clips and asked people to tell what happened in them, comparing what was said by different people, but also what was said by the same person at different times. I saw this research as an extension of Bartlett (1932), but with spoken instead of written language and with more natural stimuli. It seemed promising, and I received a grant from the National Institute of Mental Health to expand on it. I was lucky to win the collaboration of a talented group of students. We put together a short film, the so-called Pear Film, which different people took to different parts of the world, where they recorded what at least twenty people said about it in each of ten languages. In the end the project may have uncovered more about language differences than about the cognitive processes that had interested me at first, but there was nothing wrong with that (Chafe 1980).
In the early 1980s I became involved in two other projects. One stemmed from a conference on evidentiality that Johanna Nichols and I organized at Berkeley in 1981. The varied presentations at that conference led me and others to think more deeply about the ways epistemology is expressed in different languages, something that had previously received little attention (Chafe and Nichols 1986). They threw into relief an unconformity between people’s thoughts and the “real world” surrounding them, with everything we think being our own invention, even if it is constrained by our exposure to limited aspects of reality. Languages offer a variety of devices that show how people tacitly recognize and evaluate that unconformity, shedding a useful light on the extent to which people evaluate what is real and what is unreal, and how they relate to that distinction.
Before the Pear Stories project I had worked largely with languages that had seldom been written down. Most of the pear stories were oral too, but the few that were written showed striking deviations from orality. That observation led me in 1979 to organize a seminar on differences between spoken and written language. One of the participants, Jane Danielewicz, then a student in the Berkeley School of Education, became interested enough to join me in a project for which we received funding from the National Institute of Education. Realizing that spoken and written language are not just two clearly defined and separate ways of using language, she and I collected two styles of speaking and two styles of writing from each of twenty people: dinnertable conversations, talks to classes, personal letters, and academic articles. We managed to tabulate and at least partially explain a variety of ways in which spoken and written language might differ, but also, orthogonal to that distinction, ways in which unplanned language (conversations and letters) might differ from planned language (talks to classes and academic prose). We published a synopsis of our findings in Chafe and Danielewicz (1987), and Danielewicz collected parallel data from children as the basis for her dissertation (Danielewicz 1987). At first the Pear Stories project and then this spoken-written project brought home to me not only how writing might differ from speaking, but also the extent to which ordinary conversational language is the most basic of all the ways language is used, every other use being in some way a deviation from it.
At least by the early 1970s I had come to realize that language and the mind are so interrelated that it would be futile to study one without studying the other. There was a prejudice against this view that one occasionally encounters even today, a relic of Bloomfield’s dictum that “we must study people’s habits of language—the way people talk—without bothering about the mental processes that we may conceive to underlie or accompany these habits” (Bloomfield 1922, in Hockett 1970:92). Thus another respect in which I parted company from my graduate training was my wholehearted acceptance of what Bloomfield repeatedly castigated as “mentalism”.
Because they weren’t well explained by standard views of syntax, I had long been interested in the effects of what were sometimes called given and new information, and in the mid-1970s I pulled together some thoughts on that distinction and related matters (Chafe 1976). Eventually I updated and expanded that work, exploring in detail how the flow of consciousness affects the flow and shape of language (Chafe 1994). I was concerned at the same time with the important, though often overlooked distinction between immediate and displaced consciousness, immediate consciousness focused on the here and now, displaced consciousness on experiences remembered or imagined. I explored some of the ways these two modes of consciousness are expressed, both in conversations and in literature. I found what I thought were some useful ways of explaining how fiction writers manipulate them. Consciousness has lately become a hot topic in some circles, but I find that its properties and manifestations have not otherwise been investigated along the lines set forth in that book, which I regard as in some ways my most important contribution to an understanding of how language works.
Since then I have devoted more time to prosody (variations in pitch, loudness, timing, and voice quality), suggesting its place within the overall design of language (Chafe 2000a). I would like to achieve a better understanding of emotions and how they are expressed in language and even in music (obviously a large order). It is intriguing to realize that prosody is articulated primarily in the larynx, whereas segmental sounds are shaped primarily by the tongue. One is thus led to suspect the existence of special connections between the larynx and the processing of emotions in the older brain, as well as between the tongue and the processing of ideas in the neocortex. This division of labor may shed an important light on the evolution of language, not to mention the evolution of the larynx, the tongue, and indeed the brain. I lack the skills and resources to investigate neural pathways, but what I can do is search for emotions in people’s talk, trying to characterize the emotions themselves along with their prosodic manifestations. My first steps in that direction focused on the triggering and expression of excitement (Chafe in press b). Currently I am investigating laughter, a ubiquitous element of ordinary talk, which I see as the expression of a very common emotion that is seldom recognized as such. The phonetics of laughter is interesting in itself, but beyond that I am trying to sort out the varieties of experience that trigger laughter as people speak.
From time to time someone has spoken to me of what he or she has called “your theory”, a phrase that has always made me uncomfortable. I have no wish to think of myself as the promoter of some theory, preferring the role of someone who has made various suggestions derived from various things I have been able to observe, and who has tried to place them within larger frames of reference I have hoped will increase our understanding of language and the mind. I can sympathize with Dorothy Lee when she wrote:
Some years ago, I published a book of selections of my articles. I had written them for my colleagues as a form of communication. When they appeared as a book, they were used in undergraduate courses. And I found out, with horror and guilt, that the students I talked to had found the truth. I was the author, I gave the authoritative statements. I could not even argue with them because they answered me back from the authority of my book. I felt that I had dumped a load of gravel on new, thin, weak, gloriously alive grass, or even on seed that had never had a chance to sprout. (Lee 1976: 2-3; thanks to Ellen Bartee for this quote.)
It seems to me that the word “theory” is used all too loosely these days. In linguistics a theory is usually labeled a “grammar” and reduced to initials—transformational grammar (TG), relational grammar (GB), role and reference grammar (RRG), lexical-functional grammar (LFG), head-marking phrase structure grammar (HPSG), cognitive grammar (CG), and so on. Theories, or grammars, are thought to offer the advantage of being precise and testable, and there is a willingness to sacrifice richness and insight in pursuit of that goal. It is said that what makes a theory good is the fact that it can be disproved, a statement I heard repeatedly from Bloch in the 1950s. I suspect that many who adhere to this philosophy of science today are unaware of its origins in behaviorism and logical positivism, the twin philosophical foundations of American structuralist linguistics.
Unlike many who profess to be scientific, I believe that this way of “doing science” has retarded progress. Understanding is in the end an ability to place one’s inevitably limited observations within a larger and more encompassing vision. The only true way to increase our understanding is by seeking constantly to improve and extend whatever observations we can make, while at the same time improving the quality and scope of the larger vision our imaginations have created to give those observations a natural place. Confining that larger vision to some rigidly articulated theory locks us into an understanding that is inevitably deficient, simply because the forces confronting us are far too complex to be accommodated in that way. If it were the case that theories marched forward in a parade of ever-increasing progress, each an advance over what preceded, something could be said for rigid formulations at each successive stage. But the history of the human sciences, well reflected in the history of linguistics, suggests otherwise. There is no reason to suppose that a particular choice among the various competing grammars will prove to be correct, whatever that might mean. We see a collection of competing cults. I would prefer to see an opening of our understandings to continued refinement and growth as observations accumulate and improve, with a constant willingness to modify or replace our larger visions whenever necessary.
I will close with a few suggestions of a general nature. All of them assume that an improved understanding of language and the mind depends on observing as carefully as possible how language is actually used in a maximally large sample of the world’s languages. In my own work I have tried to balance observations of my own native language, where I have knowledge of a kind I could never have for another language, with the less adequate but nevertheless rewarding observations I have been able to make of other languages of very different types. But observations, as I have tried to emphasize, must at the same time be accompanied by insights that place them within a more encompassing framework. This second task is the more difficult one and the more subject to error, but one can hope to find “later views correcting earlier ones, until at last the harmony of a consistent system is reached” (James 1890:192).
Fundamental to all linguistic work should be a realization that language is a temporal phenomenon, a process that flows through time. That is partly because time is an essential ingredient of sound, but more importantly it is because thoughts flow through time as well, and language is first and foremost a way of organizing and communicating this flow of thoughts. It is futile to limit our attention to isolated sentences. The shape a sentence takes can never be appreciated without recognizing it as a small, transient slice extracted from the flow of language and thought, when it has not simply been invented to prove some point.
Because sounds are available to external observation and thoughts are not, linguists have generally tilted toward the sound side of language, mistrusting the determining role of thoughts. But it is the way we organize our thoughts that determines what we say. There is an important distinction to be made between thoughts in their entirety and thoughts as they are organized by language—a language’s semantic structuring. Each language organizes thoughts in its own unique ways whenever we speak (Slobin 1996). Sapir and Whorf were certainly right in that respect. But how much of total thought, even when we are not speaking, is affected by this language-specific semantic organization? Certainly much of our thinking consists of internal speech, and to that extent we are indeed “at the mercy of [our] language” (Sapir 1949:162). But it is the relation between linguistically organized thought and thought in its entirety that remains unsolved from the debates concerning Whorf.
I have found it important to distinguish between ideas of events and states and their participants on the one hand, and on the other hand the way these ideas are oriented—how they are placed within a complex web of space, time, epistemology, social interaction, and the ongoing talk. This distinction between ideas, so defined, and the ways they are oriented is reflected in the distinction between content words and function words or affixes. Language proceeds as a succession of activated ideas, each oriented in accordance with the semantic structuring imposed by one’s language. Interpreting experience in terms of ideas of events and states and their participants may very well be a universal property of total thought, something that exists prior to whatever ways those ideas may be categorized and oriented by a particular language during a particular act of speech. The fact that it is possible, up to a point, to translate from one language to another depends on the universality of those underlying ideas, but translations can never capture the unique richness with which each language categorizes and orients them (Chafe 2000b).
An exciting research frontier is offered by emotions and their linguistic expression. I mentioned that my own approach, one of many that are possible, has been to search recorded conversations for manifestations of affect, above all in prosody. I hope this methodology will sooner or later dovetail with brain imaging, which may provide further insights into what happens in the ebb and flow of emotions as language proceeds through time. Nor should we ignore what writers do. The success writers have lies to a large extent in their ability to manipulate the feelings of their readers, and studies of literary language can provide insights into how that happens. It would be interesting, as just one example for future research, to track neural activity as people read something sentimental like The Bridges of Madison County, or something frightening like Stephen King. How is the flow of emotion, triggered by language, reflected in the brain?
If language expresses the flow of ideas, we need to know more about ways in which ideas are linked, about the relations connecting one idea with another. One approach to this question can be found in Rhetorical Structure Theory (Mann and Thompson 1988), and I tried my hand at several other preliminary studies of discourse linkages (Chafe 1988, 1996). There is a continuing need to analyze a variety of language samples with the goal of establishing just how ideas and clusters of ideas are related as speakers proceed from one to the next. It is still not clear whether ideas are linked in a small finite number of ways that have some universal validity, or whether the possibilities are more open-ended and diverse.
At a higher level of organization, thoughts and language are formatted as a succession of topics. Each topic amounts to a partially activated cluster of knowledge within which speakers navigate with more limited, fully activated foci of consciousness. Speakers organize what they say in terms of basic-level topics, within each of which there can be a hierarchy of subtopics. At some level of this hierarchy we may encounter the kind of unit traditionally called a sentence, though I have found that sentence boundaries tend to be decided on the run as people talk, a fact which suggests that sentences may not reflect stored cognitive units of the same nature as topics and foci of consciousness. In any case the topic organization of thoughts and language needs to be studied as a fundamental determinant of discourse structure.
Turning finally to social factors that influence the shape of language, we need to keep in mind that conversation consists of interactions between separate minds and separate selves. Language is the preeminent way of compensating for the fact that our separate brains lack direct neural links. But our brains are the properties of separate selves, each with its own self-centered agenda. How communication between these separate selves is managed, both collaboratively and not-so-collaboratively, is more than a matter of taking turns. It depends to a considerable extent on the flow of topics and subtopics, but also on personal agendas, and certainly too on differing conversational styles. There is the intriguing fact that during a conversation, focal attention shifts constantly between one’s own thoughts and the thoughts of others. Different people distribute this self-centered and other-centered attention in different ways and in different proportions, with different conversational styles as a result. We need to learn more about how that takes place and the individual differences responsible for it. There is much to be learned about the obvious relation of conversational styles to personality.
These have been a few suggestions for continuing linguistic research, derived from my own work and interests. The larger goal is to achieve a better understanding of everything that makes us human, based on an awareness that language is a primary ingredient of humanness. It is exciting to realize the power of linguistics to shed light in so many ways on the richness of human experience.
Bartlett, Frederic C. 1932. Remembering: A Study in Experimental and Social Psychology. Cambridge, UK: Cambridge University Press.
Bloomfield, Leonard. 1922. Review of Edward Sapir (1921). The Classical Weekly 15.142-143.
Chafe, Wallace. 1962. “Phonetics, semantics, and language”. Language 38.335-344.
---------. 1967. Seneca Morphology and Dictionary. Smithsonian Contributions to Anthropology, vol. 4. Washington: Smithsonian Institution.
---------. 1968. “Idiomaticity as an Anomaly in the Chomskyan Paradigm”. Foundations of Language 4.109-127.
---------. 1970a. Meaning and the Structure of Language. Chicago: The University of Chicago Press.
---------. 1970b. A Semantically Based Sketch of Onondaga. International Journal of American Linguistics, Memoir 25 (Supplement to vol. 36, no. 2).
---------. 1976. “Givenness, contrastiveness, definiteness, subjects, topics, and point of view”. In Li 1976, 25-55.
---------. 1980. The Pear Stories: Cognitive, Cultural, and Linguistic Aspects of Narrative Production. Norwood, NJ: Ablex.
---------. 1988. “Linking Intonation Units in Spoken English”. In Haiman and Thompson 1988, 1-27.
---------. 1994. Discourse, Consciousness, and Time: The Flow and Displacement of Conscious Experience in Speaking and Writing. Chicago: The University of Chicago Press.
---------. 1996. “Beyond Beads on a String and Branches in a Tree”. Conceptual Structure, Discourse, and Language ed. by Adele Goldberg, 49-65. Stanford: Center for the Study of Language and Information.
---------. 1999. “Floyd Glenn Lounsbury, 1914-1998: A Brief Obituary”. Historiographia Linguistica 26:333-341.
---------. 2000a. “The Interplay of Prosodic and Segmental Sounds in the Expression of Thoughts”. Proceedings of the 23rd Annual Meeting of the Berkeley Linguistics Society, 1997, 389-401.
---------. 2000b. “Loci of Diversity and Convergence in Thought and Language”. Explorations in Linguistic Relativity ed. by Martin Pütz and Marjolijn H. Verspoor, 101-123. Amsterdam/Philadelphia: John Benjamins.
---------. In press a. “Putting Grammaticalization in its Place”. New Reflections on Grammaticalization: Proceedings of the International Symposium on Grammaticalization, 17-19 June 1999, at Potsdam University, Germany ed. by Ilse Wischer and Gabriele Diewald. Amsterdam/Philadelphia: John Benjamins.
---------. In press b. “Prosody and Emotion in a Sample of Real Speech”. Relations and Functions Within and Around Language ed. by Peter Fries, Michael Cummings, David Lockwood, and William Sprueill. London: Continuum.
--------- and Jane Danielewicz. 1987. “Properties of spoken and written language”. In Horowitz and Samuels 1987, 83-113.
--------- and Johanna Nichols. 1986. Evidentiality: The Linguistic Coding of Epistemology. Norwood, NJ: Ablex.
Danielewicz, Jane. 1987. Differences between Spoken and Written Language: A developmental analysis of adults' and children's language in fourcntexts. Ph.D. dissertation, University of California, Berkeley.
Haiman, John and Sandra A. Thompson, eds. 1988. Clause Combining in Grammar and Discourse. Amsterdam/Philadelphia: John Benjamins.
Hockett, Charles F., ed. 1970. A Leonard Bloomfield Anthology. Bloomington: Indiana University Press.
Horowitz, Rosalind and S. Jay Samuels, eds. 1987. Comprehending Oral and Written Language. San Diego: Academic Press.
Huck, Geoffrey J. and John A. Goldsmith. 1995. Ideology and Linguistic Theory: Noam Chomsky and the Deep Structure Debates. London: Routledge.
James, William. 1890. The Principles of Psychology. 2 vols. New York: Henry Holt. Reprinted 1950 by Dover Publications.
Lakoff, Robin Tolmach. 1989. “The Way We Were; or, the Real Actual Truth about Generative Semantics: A Memoir”. Journal of Pragmatics 13.939-88.
Lee, Dorothy Demetracopolou. 1976. Valuing the Self. Prospect Heights, IL: Waveland Press.
Li, Charles N., ed. 1976. Subject and Topic. New York: Academic Press.
Mandelbaum, David G., ed. 1949. The Selected Writings of Edward Sapir in Language, Culture, and Personality. Berkeley: University of California Press.
Mann, William C. and Sandra A. Thompson. 1988. “Rhetorical Structure Theory: Toward a Functional Theory of Text Organization”. Text 8.243-281.
Sapir, Edward. 1921. Language: An Introduction to the Study of Speech. New York: Harcourt, Brace.
---------. 1949. “The Status of Linguistics as a Science”. In Mandelbaum 1949, 160-166.
Slobin, Dan I. 1996. “From ‘Thought and Language’ to ‘Thinking for Speaking’”. Rethinking Linguistic Relativity ed. by John J. Gumperz and Stephen C. Levinson, 70-96. Cambridge: Cambridge University Press.