American Speech 76.2 (2001) 204-207
[Access article in PDF]
Is Linguistics a Science?
Not too long ago, I went looking for reviews of Yngve's From Grammar to Science and found that, although it has been out for about five years, it had been reviewed only once, and then only electronically on the Linguist List (Hacken 1997). Considering the book's potentially enormous impact on the field of linguistics, I was surprised, to say the least. Further, when I read Hacken's review, I was quite taken aback. It has several major errors of fact with regard to the book's contents--but more about that below.
From Grammar to Science is aimed primarily at practicing linguists. At many points, the author speaks to them directly, often using the metaphor of moving from the "old country" of a linguistics based primarily on philosophical approaches to a "new country" in which linguistics enters the domain of hard science (see especially chap. 9, "Plans for Emigrating to the New World"). [End Page 204]
The book's intent is ambitious. First of all, Yngve points out problems at the very foundation of contemporary linguistics that undermine its attempts to be "scientific." At the risk of oversimplifying, I can summarize by stating that Yngve says these problems arise essentially out of what he calls "domain confusions" between objects of the sort studied by philosophy and objects of the sort studied by science. As Yngve (like others) argues convincingly, the objects of language are more like those studied by philosophy than those studied by science. (My own corpus-based research comparing references to the research process in texts from "theoretical" and "applied" linguistics seems to confirm this [Coleman 1999].) Note Saussure's often-quoted remark on the "objects of language": "Far from it being the object that antedates the viewpoint, it is the viewpoint [of the observer] that creates the object" (1959, 8). As I have often explained it to my own students, people, the noises they make, the marks they make on paper, and so on are observable in the "real world"--what Yngve calls the "physical domain." But language (i.e., "grammar") is something linguists have offered by way of EXPLANATION for what they observe. Itkonen (1978) has shown beyond doubt the dangers to linguistics of conflating the former sense of "observation" (recording sense data from the real world) with the one in which we speak of "making observations" about hypothetical entities. People, Yngve argues, are objects in the physical domain; language is not. "We cannot found a science on an assumption that creates its own objects of study," that is, "objects of language" (33). The linguistics he proposes is, rather, "a human linguistics, a linguistics focused on people rather than on language" (121). He takes most of chapters 1-9 to fully develop his arguments along these lines.
Yngve outlines his conceptual framework in chapters 10-18. He shows how a system of attaching properties to individuals and interactional groups in which they participate can be developed into a complex apparatus for describing human communication. It is disappointing but perhaps understandable that he provides only short examples of various communicative situations and does not develop any of his examples in great depth. The nature of his framework--which includes properties of individuals participating in a communicative event, properties of the interaction, and descriptions of conditional changes of state in these properties--would require a few chapters to develop an in-depth example, and would require prior understanding of the conceptual framework he has developed.
In chapters 19 and 20, the author presents a notational system that he argues will allow linguistics to be truly scientific. To this end, he does not [End Page 205] create a new notation out of thin air, but adopts standard diagramming conventions from Boolean logic.
Now, here are some of the blatant misrepresentations made by Hacken (1997). "In Yngve's view scientific knowledge is permanent and represents truth," a position Hacken describes as "so naive no philosopher would get...