TRUE CONFESSIONS? PRAGMATIC COMPETENCE AND CRIMINAL CONFESSION
Robin Tolmach Lakoff University of California, Berkeley
There are two especially difficult problems for pragmatics and sociolinguistics. One is the connection between purely linguistic form (phonology, syntax, morphology) and discourse function. Eventually we must, however reluctantly, consider to what degree and in what ways real-world situations and communicative needs of speakers govern syntactic form; and the sense in which syntax, semantics, and pragmatics are truly, bidirectionally, interdependent. The second is the definition of "pragmatic" or "communicative" competence: What does the normally competent speaker know?
The case I will discuss here raises another issue of interest to linguistics and other social sciences. Over the last fifteen years or so, there has been much discussion in several fields concerning the nature and/or reality, of the "self." Some of the data we will be examining provide more evidence of the tenuousness and fuzziness of that concept.
The examples I will use are drawn from a pair of criminal confessions in a death penalty case currently ( June, 1993) under appellate review. I was asked as an expert by the defense to review the confessions to assess the linguistic (or communicative) capacities of their client. Since I was asked by counsel to use a pseudonym in referring to the defendant, I shall refer to him as Virgil Reilly (or VR).
The facts of the case are briefly these. When the crimes were committed, in November, 1984, Virgil Reilly was a black male of 25, who had never been in any previous legal trouble.
VR's IQ falls between the mid-60s and mid-70s -- borderline or "dull normal." Psychological testing reveals an array of impairments -- particularly relevant to our concerns are verbal deficits, including "severe impairment in attention, memory for verbally presented material, verbal fluency, cognitive flexibility, and adaptability to novel