Language An Introduction to the Study of Speech

by Edward Sapir

1921. This little book aims to give a certain perspective on the subject of language rather than to assemble facts about it. It has little to say of the ultimate psychological basis of speech and gives only enough of the actual descriptive or historical facts of particular languages to illustrate principles. Its main purpose is to show what I conceive language to be, what is its variability in place and time, and what are its relations to other fundamental human interests-the problem of thought, the nature of the historical process, race, culture, art. Contents: Language Define; The Elements of Speech; The Sounds of Language; Form in Language; Grammatical Processes; Form in Language; Grammatical Concepts; Types of Linguistic Structure; Language as a Historical Product: Drift; Language as a Historical Product: Phonetic Law; How Languages Influence Each Other; Language, Race and Culture; and Language and Literature.

Genres: Linguistics, Language, Nonfiction, Philosophy, Anthropology, Science, Psychology

110 pages, Kindle Edition

First published January 1, 1921

About the author

Edward Sapir was born in Pomerania, Germany, in 1884, and came to the United States at the age of five. He first made his reputation as an expert on languages of the Native American. He taught at the University of Chicago and later at Yale, and was one of the first to explore the relations between language studies and anthropology. He died in 1939. Language, first published in 1921, is his only full-length book for a general audience. He published a great many articles and some verse in periodicals.

Reviews

[Carl · Rating 3 out of 5]

I am watching John McWhorter’s Great Course on Linguistics, in which he recommends Sapir’s book as a general introduction to linguistics. After reading the book I thought that it was a good one to read before you start studying linguistics – it seems to cover a lot of the field, giving the reader a feel for the subject. But the books neglects to define some of the basic technical terms that a beginner needs, so that it also seems like a summary aimed at a scholar who has spent several years in the field.

The book is about 90 years old, so I don’t know how much has changed in the field since then, but McWhorter recommended it, so I suppose it must still be current.

Sapir’s erudition is colossal. He gives examples from languages from all over the world, and of many types. He really seems to know what he is talking about.

Something that I thought was amusing was Sapir’s discussion of the word “whom”. He spent several pages on it. He thought “whom” was on its way out of the English language, citing several tendencies in English that worked against it. Now, 90 years later, John McWhorter also sneers at the word. It is true that we often feel awkward with “whom”, and often aren’t quite sure when to use it or not, but it just seems to keep tottering on and on. I wonder why.

[Austin Lynch · Rating 4 out of 5]

4.5 stars. Very interesting survey of some fundamentals of linguistics (grammar, phonology, drift), especially as related to English specifically. I'm not a linguist and can't speak to how well this 1921 book holds up to modern linguistic theory, but it's highly readable and well structured.

There's an excellent ebook with working endnotes (this book contains many endnotes and all are worth reading) available for free on Project Gutenberg.

[Sergio · Rating 4 out of 5]

I'd give it a higher rating, but the sections on phonetic law and its various classifications went over my head. However, the chapters that were more grounded in sociology and aesthetics were fantastic. Still, as a layman I found Sapir's writing very accessible despite the technical sections. I'm guessing some of this stuff is outdated, but it was definitely worth reading.

[Gary Bruff · Rating 5 out of 5]

Edward Sapir. Language: An Introduction to the Study of Speech. 1921.
Kilu von Prince, Ana Krajinović, and Manfred Krifka. “Irrealis is Real”. 2022.

When I was an undergrad studying anthropology back in the mid-1980’s, Sapir’s Language was one of the first books that really spoke to me. It seemed to embody what anthropology was all about. While most of today’s mainstream linguists have taken to criticizing or ignoring Sapir’s methods and discoveries of a century ago (for example, see Steven Pinker’s extended diatribe against Sapir in The Language Instinct), Sapir for me still occupies a special place in the canons of both linguistics and anthropology. In 1921, Sapir thought of the young discipline of linguistics as a real science, yet many of the metaphors and images he uses to characterize LANGUAGE in the abstract appear to be anything but scientific. But I maintain that Sapir is nevertheless consistent in his approach to language. If he flexes his poetic muscles at times, it is to show how language gives its speakers the power to stretch the limits of their understanding, allowing them to say what has never before been said, and to mean it.

Some of Sapir’s metaphors about language bear repeating. One group of metaphors link together thought and language:

“...thought is nothing but language denuded of its outward garb.”
“...language, as a structure, is on its inner face the mold of thought.”
“...speech would seem to be the only road we know of that leads to [thought].”
“...the word, as we know, is not only a key; it may also be a fetter.”
“Culture may be defined as what a society does and thinks. Language is the particular how of thought.”

Another group of metaphors depicts language as a work of art:

“Single Algonquin words are like tiny imagist poems.”
“Every language is itself a collective art of expression.”
“Language is the most massive and inclusive art we know, a mountainous and anonymous work of unconscious generations.”
“Language is the medium of literature as marble or bronze or clay are the materials of the sculptor.”

These images I hope demonstrate that Sapir practiced what he set out to preach. In contradistinction to Chomskyite formal methods and mathematical modeling, Sapir’s tropes give free reign to the power of words to say what has not been said before and to express what possibly cannot be expressed otherwise. From the point of view of Chomsky's Universal Grammar (sub specie aeternatis, as Spinoza might say), Sapir’s linguistic typology does violence to the routine generativist assumption that all languages are underlyingly identical in structure. For Chomsky, and for all other linguists who know not much about languages besides English, language is not a work of art, nor is it the garb of thought. It is the computational system of human cognition, bridging universal and innate ideas to various means of articulating those ideas. But then again, metaphor is at work here, too.

Sapir did not conjure his typological and historicist approach to linguistics out of whole cloth. Rather, he derived many of his ideas about language from German philologists like Wilhelm von Humboldt. But unlike Humboldt, Sapir knew that factors like race, national temperament, or even level of cultural complexity can in no way be tied to the ‘genius’ of a given tongue. Sapir gives the example of English as spoken by African Americans:

“The English language is not spoken by a unified race. In the United States, there are several millions of Negroes who know no other language. It is their mother-tongue, the formal vesture of their inmost thoughts and sentiments. It is as much their property, as inalienably ‘theirs,’ as the King of England’s.”

Unlike Chomsky, and along with Humboldt this time, Sapir divides speech into two key aspects: language as creative activity (energeia, or language as the ‘key’) and language constrained by the habituated sediment of past expression (ergon, or language as the ‘fetter’). Speakers in pursuit of meaning can at times take an active role in fabricating their language, for “language is a purely human and non-instinctive method of communicating ideas, emotions, and desires by means of a system of voluntarily produced symbols.” But the historical context of our language inevitably plays a role in delimiting our speech as well, channeling our thoughts into the cognitive ruts left behind by all that has been said before by speakers of our language.

The fact that each language is unique does not contradict Sapir’s creed that all languages are equivalent, that is, that they are all equally valid. Whereas no language is perfect, all languages are in fact “fully developed”, even though they reflect “an almost incredible diversity.” All languages moreover demonstrate “a certain randomness of association,” which means that our semantic categories may lump together certain ideas while splitting other ideas apart. (Some facile examples from the simpler domain of vocabulary: Unlike English, Chinese makes a lexical distinction between rice as plant (mi) and rice as food (fan). English meanwhile distinguishes goats from sheep, while in Chinese they are both called yang.) There is a tension between our need at times to break grammatical rules and the ordinary obligation to follow those rules, and the apparent imperfections of grammatical systems seem to arise from this tension. Grammars then are not innate but are in fact nearly always a work in progress. As such, grammatical systems are forever incomplete: “Unfortunately, or luckily, no language is tyrannically consistent. All grammars leak.”

If there is one sub-field of linguistics that still finds favor with Sapir’s methods and insights it would have to be typology. Sapir’s own very dated typological system, presented in the “Grammatical Form” chapters of Language, is not really all that useful to the modern student. Despite its name, typology these days is not about typing languages into neat pigeon holes in order to make a final classification of all languages. Instead, typologists seek to discover a useful metalanguage by an extensive inventorying of grammatical categories, categories which are found in some languages and not in others. Empirically, this is what makes an honest comparison of grammars possible. Unlike generativists, typologists do not regard the seemingly endless variety of grammatical systems as a source of consternation or embarrassment. While followers of Saint Noam may be convinced that the empirical diversity of grammars will one day very soon be reduced to rigorous order and timeless uniformity, typologists willingly embrace that same empirical diversity, keeping their minds open to the astonishingly different ways in which grammars can be said to vary. As Sapir put it:

“...if we take our examples freely from the vast storehouse of language, from languages that are exotic as well as those that we are more familiar with, we shall find that there is hardly a possibility that is not realized in actual usage.”

A century after Sapir’s Language appeared, many linguists of a typological persuasion are still dredging up to the surface the kind of grammatical categories which are prominent in some languages but which are absent from others. One such recent study, “Irrealis is Real” by von Prince, Krajinović, and Krifka, looks at the categorical contrast of realis and irrealis. As the article points out, Sapir had already described an irrealis suffix in his 1930 grammar of Southern Paiute: “This element indicates that the activity expressed by the verb is unreal, i.e. either merely potential or contrary to fact (potential in past time)” (Sapir: “Southern Paiute, a Shoshone Language”). In many of the Oceanic languages of the South Pacific, there is this same obligatory grammatical distinction, part of the tense-aspect-mood system of their verb paradigms, by which speakers categorize all events as either actual or else as non-actual. So irrealis, marking the non-actual, is used to indicate future events, hypothetical possibilities, or counterfactual scenarios. All languages can make these kinds of semantic distinctions, but languages with irrealis have grammaticalized a complex idea onto a single morpheme for indicating a modal and often temporal contrast between the actual and the non-actual.

The authors of “Irrealis is Real” identified in their typological studies of Oceanic semantics some of the typical uses of irrealis verb markings: “irrealis markers are used for talking about future and counterfactual events. They are used in future and counterfactual conditionals, in complement clauses of verbs expressing wishes and intentions, in purpose clauses, and in expressions of ability and obligation.” It might be best to think of irrealis in model theoretic terms; irrealis serves to indicate a possible scenario (possible world) which the speaker judges to be incongruous with what we know with certainty to be the truth of the actual scenario, either because something hasn’t yet happened or else because some event is merely desired or speculated upon. In many European languages, the subjunctive mood plays a somewhat similar role to irrealis in being frequently merged with future tense, in bearing a hypothetical/wish connotation, and also by being used in subordinate clauses. Taking a typological approach to grammar, we can make use of ‘exotic’ languages in order to see more familiar languages in a new light.

But typologically speaking, the important thing to remember about irrealis as a unified semantic category is precisely the fact that this form of grammatical contrast is not universal. Rather, reallis/irrealis appears to be localized to some languages in certain places, like the South Pacific. In a sense, irrealis is like other grammatical categories which appear in some languages but not in most—for example, definiteness (many languages do not have articles) and tone (you oftentimes find tone languages in Bantu Africa and East Asia, but not much elsewhere).

It is perhaps fitting to end this review of Sapir’s Language by asserting once again that most rationalist assumptions of formal uniformity in all grammars are reductionist. What’s preferable is an empirical method that shows we are all human, not because we all think the exact same thoughts, but because humans at different times and in different places had arrived at the same solution for understanding reality (and in this case, irreality). Having merged under a single cognition ideas that most languages keep distinct, the South Seas islander and the Paiute Native American came to share a grammatical category of irrealis--despite being separated by 8,000 miles of ocean, desert, and the Grand Canyon. I find this patterning of variation to be interesting, but not altogether surprising. And I find it somehow reassuring that all languages are in fact potentially the same in their choice of semantically restrictive categories. But that potential does not imply that two languages separated by time and distance must somehow be cognitively identical. As any good anthropological linguist would say, ‘Vive la différence!’

Sapir’s Language is in the public domain. You can get it in lots of places, including gutenberg.org. “Irrealis is Real” was published in the journal Language vol 98:2. Paul Friedrich’s book The Language Parallax reflects on the poetic spirit of both Sapir and of Native American speakers. Rhetoric and Grammar, my master’s thesis at Rice, was written in part about Sapir’s rhetorical and grammatical contributions to linguistics and the other human sciences. A good recent reference for typology can be found in Payne’s Describing Morphosyntax and in Mithun's Languages of Native North America.

[James F · Rating 4 out of 5]

I said in my previous review of Lévi-Strauss' Myth and Meaning that I was going to read five secondary works on Structuralism next; I had hardly begun reading the introduction to the first book before I realized that I needed to read a few other books on structural linguistics, philosophy, psychology and literary theory first, as background. By a few, I mean about twenty. This is the kind of regress that keeps me from ever getting through my planned reading lists. To be sure, most of them were already on my extended TBR list; some, like Derrida and Lacan, were authors I have deliberately procrastinated on reading, others like Sapir were authors I actually wanted to read.

Sapir was one of the early pioneers of structural linguistics; Language, published in 1921, is one step beyond de Saussure, the founder, whom I read a few years ago, but well before my college linguistics textbook from the late sixties. Naturally, being just over a hundred years old, it is in some respects outdated, especially in the more technical parts, but the linguistic scientists who have superseded him all built on his foundations; that's how science works. For the most part, however, the book is fairly non-technical and is still a good general introduction to language for the general reader, and of course a must-read for any student of linguistics who is interested in the historical development of the discipline.

Sapir disclaims any attempt at dealing with the neurological bases of language; he goes a bit too far when he says that language merely uses abilities developed for other purposes in the same way as we use our brains to do math or our fingers for playing the piano — we certainly know today that areas of the brain as well as the vocal apparatus have evolved specifically to use language. However, what he was getting at is that languages need to be studied as abstract structures, which is true. In his discussion of phonetics (chapter 3), he approaches the idea of the phoneme, without using the term or defining it as clearly as contemporary linguistics books do. His discussion of the basic processes and concepts of grammar and how they are implemented in such diverse ways by different languages (chapters 4 and 5) is the heart of the book. The chapter on types of languages (chapter 6) is the most difficult, because he is trying to work out a new basis of classifying languages; the general reader may want to skip or skim through this section, although I found his discussion of agglutination vs. fusion very interesting.

Chapters 7 through 9 deal with diachronic linguistics (language change) and are the best thing I have ever read on the general principles involved, even though much has been learned since in detail. His example of the (lingering) disappearance of the "who"/"whom" distinction in English, and his other examples of the direction in which English is changing (e.g., "It's me" vs. "It is I") are very insightful and written in a lively, almost humorous style, and the history of the language over the last century has borne out many of his predictions. One phenomenon he did not predict, of course, was the contemporary usage of "they" and "their" as gender-indeterminate singulars in sentences such as "If someone has a toothache, they should see their dentist" — okay, I'm not great at thinking up examples — which I do not think is just an "incorrect" failure of agreement between singular antecedent and plural consequent as prescriptive English teachers claim, but an actual use of a former plural as a singular, completely analagous to the sixteenth-century use of the formerly plural "you" and "your" in place of the singular "thou", "thee" and "thy". Nevertheless, I think it fits well into his overall argument.

The blurb on the back of my edition (a Dover reprint) mentions the controversial "Sapir-Whorf Hypothesis", which claimed that at least some concepts are so tied to specific languages that they cannot be understood by speakers of other languages. There is nothing resembling this in the current book; he and Whorf developed that idea later, and in fact Sapir seems to say just the opposite in discussing the relationship of language to culture in chapter 10: "Nor can I believe that culture and language are in any true sense causally related. Culture may be defined as what a society does and thinks. Language is a particular how of thought. it is difficult to see what particular causal relations may be expected to subsist between a selected inventory of experience . . . and the particular manner in which the society expresses all experience." The final chapter does however make the much less questionable point that the structure of a particular language (e.g., whether it has a fixed word order or a more variable one, whether it has inflectional endings, whether it can freely form compounds, etc.) may determine what sort of literary styles can be successfully developed in that language.

[Adrian Peters · Rating 5 out of 5]

Language: An Introduction to the Study of Speech by Edward Sapir.
I read this book years ago and just recently picked it up again and began to read it again. It really is a classic in the field of linguistics - the scientific study of language. Written almost a hundred years ago it still rings true and remains relevant to contemporary approaches to language. This is more than one can say of Chomsky's claims for a language faculty or a universal grammar back in the 70's. I guess the best we can say of that wild goose chase was that it was a good try but failed to provide any evidence to support its claims. Even the claim that recursion exists in every language has come under criticism.
Sapir begins by comparing talking to walking and concludes that whereas we are bound to stand up and begin to walk the same is not true of speech which depends upon the circumstance of being born into a society. He writes how language 'varies without assignable limits' and how it is 'a purely historical heritage' 'the product of long-continued social usage'.
In the same first chapter, he goes on to compare language to thought. No mean feat in itself, he deals first with the localisation of language in certain areas of the brain separating its electro-neurological activity from language on the grounds that it has not yet assumed a 'meaning'. With superb clarity, he concludes by citing meaning as the hallmark that identifies the linguistic experience.
All of which makes me wonder why so many - Chomskyites, Pinker, Fodor etc - find the diversity exhibited by 5,000-odd languages so unpalatable? The objective of science is to describe laws that explain the various states of all phenomena. Language appears to be one that has eluded that definitive form of description.
As an offshoot of human consciousness, is it any surprise that language slips through the net of science? It seems to me that the problem we face today and ever more so in the future is how technological developments will use language for deception - to influence elections, referenda, opinion, minds - until it can no longer be trusted and someone will write: In the end, the Word was Bad.

[Dave · Rating 4 out of 5]

This book is a fantastic introduction to the complexities of human language and gives a good overview of the various categories of study that linguistics has been divided into (e.g., phonetics and morphology). Sapir's particular attention to the framework and intrinsic "genius" of each language helped me appreciate how distinct and different languages could be and temporarily escape the simple English-language mindset that I typically approach the topic with.

All of that said, if an "introduction" was indeed Sapir's goal, he was a bit off the mark. Some of the material is quite dense and presupposes a basic understanding of linguistics that the book itself purports to offer. It would probably need to be twice as long to provide that basic context and still reach as deep as it does into the material.

[Hyang Magallanes · Rating 4 out of 5]

Tiene dos Capitulos en medio que son muy cansinos porque son muy insistentes para probar un punto. Pero pues, era 1920, no podemos pedir mucho.

Pero en general, es un libro casi obligatorio para cualquiera interesado en lenguaje y filosofía.

[Consuelo · Rating 4 out of 5]

Read Lenguaje, Raza y Cultura.

[Ash · Rating 5 out of 5]

Dated, and while full of wild claims (that nevertheless were not so wild at the time of writing), I would consider this essential reading.

[Melrose · Rating 5 out of 5]

This book is a well said for amyone beginning a linguistic view for the shelf life

[Hailey Hartman · Rating 5 out of 5]

Although some of the more technical aspects of this book flew miles above my head, it was an incredibly worthwhile and interesting book.

[Milo Giménez · Rating 3 out of 5]

Un tema extremadamente interesante explicado de la manera más aburrida posible

[Madelyn Weber · Rating 2 out of 5]

1. Ejemplos confusos y no pertinentes.
2. Léxico muy rebuscado y poco claro.
3. Secuencia de los hechos rebuscada y difícil de seguir.

[Khurin]

https://www.gutenberg.org/cache/epub/...

[baal · Rating 5 out of 5]

It is rather easy to dismiss what E. Sapir started in the study of linguistics if we only understand this science in the aftermath of L. Bloomfield or N. Chomsky; however, Sapir's proposal for a new perspective on grammatical processes underlies every single typological and morpho–syntactical theory nowadays.

[Matthew · Rating 3 out of 5]

Deserves a lot of respect for being one of the pioneering works in the field of linguistics, but all the information it contains is available in much more readable works that have been published since. For linguistics majors only.

[Charles]

I'm not giving this book a rating because I'm in no way qualified to judge it. It is a very technical book that will appeal to people who like to look at structures and formulas and find that a helpful way to look at the world. For me it was utterly exhausting, and I gave up fairly quickly.

[Mark Colombo · Rating 2 out of 5]

This book may have been very good for what it is, but I had a lot of trouble getting through even the first chapter and ultimately didn't finish it. This book isn't really a "read for pleasure" type of book, or at least, it wasn't to me.

[C. Todd White · Rating 3 out of 5]

This was a good book but not a great one. While the word is that this is a good primer to linguistics, I found it to be cumbersome where it most needed clarity. It was mercifully playful in some areas, though, reflecting on language as art.

[Don · Rating 5 out of 5]

A fantastic short, deep plunge into the nature of language and speech. A wonderful intro to the field of linguistics.

[Eder · Rating 3 out of 5]

Obviamente, en español.

[Nazila · Rating 4 out of 5]

I not only read this book but also revised the translation. Its revision has a long history. It was my prof who translated it and asked me to revise it.

[David Hildebrand · Rating 3 out of 5]

It's historically interesting and surprisingly modern.

[James Tanner · Rating 5 out of 5]

This is the must read introduction to linguistics

[Des Bladet · Rating 5 out of 5]

Still my favourite introduction to linguistics.