Descriptive linguistics is a subfield of linguistics that studies and describes language in structural terms. In its investigation of linguistic structure, descriptive linguistics emphasizes the primacy of speech, the adoption of a synchronic approach, and the description of language and dialect systems as they are found to be spoken.
Descriptive Linguistics as a Scientific Pursuit
The rise of descriptive linguistics is generally attributed to the work of Ferdinand de Saussure (1857-1913), a Swiss linguist who is credited as being the father of modern linguistics. In the Cours de linguistique générale, a collation of his lecture notes published posthumously in 1916, Saussure laid out the general principles and methods of what has come to be known as descriptive linguistics.
The primacy of speech. Fundamental to Saussure’s teachings was the dichotomy of parole (speech) and langue (language). Language is a system of rules and conventions shared by a community of speakers, whereas speech is individual and particular utterances. The descriptive linguist uses speech to construct the underlying common structure or system of a language. In the construction of the language system, speech is primary and writing is secondary. Not only is speech older, but all writing systems are based on units of speech, yet no writing system captures all of the information that is carried in speech by variations of pitch, stress, gesture, facial expressions, space manipulation, context, and so on. There is also an overwhelming majority of languages that do not have a writing system. The primacy of speech, however, does not negate the importance of writing where homophony makes speech ambiguous, writing is completely independent of speech (Latin), or speech and writing have diverged significantly (Cantonese, Mandarin).
Synchronic approach. Descriptive linguistics is also known as synchronic linguistics. The synchronic study of a language involves describing it at a particular point in time without recourse to historical information. In contrast, the diachronic study of a language involves describing its historical development through time. Diachronic description is based on the knowledge of previous synchronic states through which languages have passed historically. By necessity, synchronic description (descriptive linguistics) is prior to diachronic description (historical linguistics). This primacy of synchronic description also implies that diachronic information is irrelevant to the investigation of a language in a temporal state. To illustrate this point, Saussure used a famous analogy that compared languages to games of chess. The pieces of a chess set may be made of wood, ivory, onyx, metal, or something else. A treatise on the origins of the chess pieces or on the history of materials used to make them may be interesting, but it has no bearing on the understanding of chess games. What is crucial for an understanding of chess games is to know the synchronic relations among chess pieces. Equipped with this knowledge, the state of the chessboard at any particular time can be described without recourse to the previous combination of moves that has brought the game to that point. Similarly, any language can be described without recourse to information about its previous states. Unlike historical linguistics, descriptive linguistics is interested in all languages, not just those languages whose historical changes are known.
Descriptivism versus prescriptivism. Descriptivism is the policy of describing a language as it is found to be spoken. Descriptive linguists write rules and grammars to capture the knowledge that speakers have of their language in producing speech. The end product is a descriptive grammar, a model about how people speak and what they know of their language unconsciously. For Saussure, the goal of linguistic analysis is to uncover out-of-awareness native knowledge, a langue as opposed to parole. Prescriptivism is the belief that people must conform to the rules laid down by prescriptive grammarians and that deviation is corruption. A prescriptive grammar dictates how people should speak. It is not unusual that prescriptivists condemn utterances that appear to be perfectly normal to the majority of speakers, for example, “It’s me” instead of the prescribed form “It’s I.” Language change is the worst enemy of prescriptivism.
Structuralism. Saussure saw language as “a system of relations.” To analyze the structural importance of relations, he developed a number of dichotomies, including the famous syntagmatic-paradigmatic dichotomy. Syntagmatic relations are contracted between units that can co-occur, whereas paradigmatic relations are contracted between units that may substitute for each other. The Saussurean point that linguistic units have no validity independent of their syntagmatic and paradigmatic relations with other units is well taken. With contributions like this, Saussure made the structuralist approach explicit for the study of language. It was Leonard Bloomfield (1887-1949) who propounded a general theory of language structure. Not only did his Language contain a thorough theoretical excursion, but the 1933 book also mapped out a detailed methodology for analyzing all levels of linguistic structure up to syntax. At the heart of Bloomfield’s theory is the autonomy of language, a thesis stating that the patterns of language operate with regularity and constitute an object of scientific study in its own right. Edward Sapir (1884-1939) and Bloomfield were the first two leading linguists of American structuralism to emphasize the importance of a descriptive approach to the study of language. They developed a substantial agenda for the autonomous, synchronic, and structural description of languages.
The goal of descriptive linguistics is to “discover” a grammar by performing a set of operations on a corpus of data. The levels of grammatical description are to be arrived at in the following order:
- Phonetics (speech sounds)
- Phonemics (phonemes and allophones)
- Morphemics (morphemes and allomorphs)
- Syntax (sentence construction)
- Discourse (use of speech)
This discovery procedure typically leads to the generation of three descriptive studies: a grammar, a dictionary, and an edited volume of texts in the language.
Descriptive linguistics features a rigorous application of basic concepts and analytical procedures. The basic units of analysis (for example, phone, allophone, phoneme, morpheme, allomorph, immediate constituent) are carefully defined, and the analytical procedures or modi operandi are laid out. The resultant descriptions are characterized by a systematic exposition of the phonemes of the language, their phonetic variants (allophones), and their realization in the morphemes (morphonemics). This is followed by a treatment of the morphology and syntax of the language that is no less rigorous and systematic. By focusing on the shared structural features cross-linguistically, descriptive linguistics enables the classification of the world’s languages typologically.
There is also a procedural constraint against “mixing levels.” As was shown earlier, in terms of analytical levels, phonemics precedes morphemics. Therefore, morphemes are to be discovered only after the phonemes of which they are composed are isolated from the data of speech. Moreover, morphemic or syntactic information must not enter a phonemic description. All of this is designed to ensure the scientific rigor of linguistic analysis against circularity. If two fairly bright graduate students who are trained in descriptive linguistics at two different institutions are sent out to describe an unanalyzed language at two different times and using two different informants, chances are that their analyses would be highly similar.
What has been presented is a “working up” model. It starts with the identification of speech sounds and phonemes and goes progressively up to examine how they are arranged into meaningful units (for example, morphemes, words, phrases) and what rules are used to combine these units into longer sentences and utterances. Conversely, the larger units of speech can be taken as the starting point as in the generative tradition. Sentences, for instance, are elicited directly and then broken down progressively. This “working down” model, however, is not what ethnographers usually follow in fieldwork. Because ethnographers must learn to speak the native language for participant observation, they tend to work up, beginning with the minutiae of the language, uncovering its structure bit by bit, and in the process managing to use its rule systems and everyday speech as a valuable source of insight into the culture of that speech community.
Descriptive Linguistics and Anthropology
Descriptive linguistics received its formative impact from anthropology as much as from linguistics. It was Franz Boas (1858-1942), the father of American anthropology, who set the stage for the development of a modern linguistic science in the United States. He developed a scheme for the systematic description of languages and outlined it in the introduction to the Handbook of American Indian Languages (1911-1922). The scope of description envisaged by Boas was to include the psychology of meaning expression in addition to linguistic structure. In 1917, he founded the International Journal of American Linguistics, which was devoted to the descriptive study of the American Indian languages.
The history of American linguistics since Boas’s groundbreaking impact is the history of the influence of Sapir and Bloomfield. As a student of Boas at Columbia University, Sapir was trained in anthropology. Like Bloomfield, he switched to the study of synchronic linguistics after doing brilliant work in historical linguistics. Whereas Bloomfield focused on the Algonquian languages around the Great Lakes, Sapir’s principal work was with the native languages of the western United States and Central America such as Kwakiutl, Chinoook, Yana, Wishram, Wasco, Ute, Nootka, Nahuatl, Chenyenne, Yurok, Comox, and Navajo. Sapir was perhaps the greatest field-worker in the history of linguistics. The analyses of a number of American Indian languages that he presented have remained valid to this day. Especially noteworthy is his Takelma grammar of 1911 (published in 1922), in which he worked out the basic principles of structuralism before Saussure’s Cours was published.
During the 1930s and 1940s, one of the primary concerns of descriptive linguistics was phonemics, which involves the classification of phonemes. Consider the p-like sounds in the words pray and spray, which are grouped under the same phoneme p in English. Acoustically, however, these p-like sounds are different, with one being aspirated and the other being unaspirated. Although this variation of aspiration is ignored phonemically in English, it is important in languages such as Sandhi because it gives rise to two different phonemes: an aspirated ph and an unaspirated p that differentiate meaning in words, as in pTMnu (snake food) and pTMnu (leaf). Conversely, where English speakers distinguish between the sounds p and b, as in peach and beach, speakers of Arabic do not and group p and b under the same phoneme. The list of examples can go on and on.
Difficulties arose when efforts to provide an object basis for defining the phonemes of an unknown language in face of their variation in sounds were attempted. A breakthrough came when Sapir proposed that the phoneme was a psychological unit under which the native mind grouped a number of similar speech sounds (allophones). It was shown that the grouping of allophones into distinctive phonemes varied from language to language. Therefore, one could not impose the forms of one’s own language on other tongues. Rather, each language had to be studied in its own terms and categories. Principles like these enabled the descriptive linguist to overcome the ethnocentrism to which Boas had objected.
Sapir also believed that linguistic structure had an underlying mental reality. To put it in Boas’s words, language was the window of the mind. This belief was reinforced by the discovery that both the phoneme and morpheme were psychologically real and could not be explained away merely in behavioral terms. It led Sapir to assume that linguistic structure also played a role in shaping people’s perception of reality, an idea that was mentalist and further developed by his student Benjamin Lee Whorf (1897-1941).
Bloomfield instead insisted on an empiricist orientation that allowed only statements based on direct observation or generalizations that were derived from observables. During his later years, Bloomfield was known for his stark refusal to consider semantics or mentalism as a part of linguistic science. It was his belief that neither could be studied scientifically. Under the predominant influence of Bloomfield’s empiricism, descriptive linguistics came to focus on the internal structure of language without reference to meaning or psychology.
The descriptive linguist tends to rely on “direct elicitation” in data collection. To the extent that direct elicitation requires a native informant to reflect on language context free to a nonnative speaker (the field-worker or researcher), biases are inherent in the data collected. Some of the biases arise from decontextualized unnatural uses of language in elicitation sessions. Others are due to the fact that the informants may use simplified speech, such as “foreign talk,” to cope with the abstract unnatural tasks imposed on them. Participant observation provides a remedy in that regard by helping the field-worker to observe native speech in context and to relate it to the meaning system and cognitive patterns of the native culture.
Descriptive linguistics has proved to be productive in the hands of linguistic and cultural anthropologists who retain the mentalist outlook of Sapir. Led by Morris Swadesh, Stanley Newman, Mary Haas, and C. F. Voegelin, they are responsible for the many careful descriptive studies of American Indian and other indigenous languages that have been published and continue to be published. It is important to include the study of linguistics in the basic training of anthropologists. The insight that structural linguistics provides into the rule and meaning systems of language is indispensable for anthropologists attempting to examine language in broad terms.
- Bloomfield, L. (1933). Language. New York: Henry Holt.
- Carroll, J. B. (1955). The study of language. Cambridge, MA: Harvard University Press.
- Dinneen, F. P. (1967). An introduction to general linguistics. New York: Holt, Rinehart & Winston.
- Duranti, A. (1997). Linguistic anthropology. New York: Cambridge University Press.
- Lehmann, W. P. (1972). Descriptive linguistics: An introduction. New York: Random House.
- Lyons, J. (1968). Introduction to theoretical linguistics. New York: Cambridge University Press.
- Newmeyer, F. J. (1986). Linguistic theory in America. San Diego: Academic Press.
- Sapir, E. (1949). Selected writings of Edward Sapir in language, culture, and personality (G. Mandelbaum, Ed.). Berkeley: University of California Press.