Conversation as Paranoia—Hugo Aelgh Lingua Pranca Contents An Iñupik Linguistic Fragment (or, the Last Grammarian)—Metalleus

Natural Phonological Processes in Adult and Child Speech

David Ingram
University of British Columbia

In this brief note, I would like to lay out some of the theoretical issues involved in the examination of phonological processes in synchronic language study and the search for these processes in children’s speech. First, the nature of phonological processes (PPs) in children’s speech (CS) has been pursued in several works, especially Stampe (1969). PPs in adult synchronic languages (AL), on the other hand, are spread across a variety of studies, both taxonomic (TP) and generative (GP), although they have been a major focus of Stampe’s theory of natural phonology (SNP) and the version discussed as natural generative phonology (NGP) as in Hooper (1976). The point is then this: are PPs in CS isomorphic with those found in AL independent of documentation in TP, GP and SNP? Or, more formally stated:

(1) {x/PP(x) ∧ CS(x)} ≠ {x/PP(x) ∧ AL(x)}

(1) constitutes the Ho (null hypothesis) on PPs.

One way to pursue this issue is to take a PP and examine its occurrence in both CS and AL. Below are a few PPs with reasonable documentation:

(2) palatalization of apicals (PA)

(3) stopping of fricatives (SF)

(4) loss of unstressed syllables (LUS)

(5) lenition of consonants (LC)

There are numerous straightforward possibilities: we might find a reasonable set of data for a PP like LUS in CS but not in AL. Or, suppose the overlap is within a particular theory, so that all PPs occur in NGP but not GP, or that some sub-class such as SF and LC has such characteristics. C & H in SPE, for instance, discuss some possible PP that, based on NGP, would not exist in AL, and presumably not is CS. This, however, raises the separate issue of possible PP (PPP) for any language and APRPPs (actual psychologically real PPs). It may well be that a child Cx learning PPP has little evidence from AL to construct an APRPP for Lw that the child is learning in LAP (the Language Acquisition Process). Thus, for any particular language, Ly, a PP may be a PPP, but due to the constraints on the LAP by the LAD, as specified let us say by our PT (phonological theory) such as NGP, the PP in question is not an APRPP.

The issue then is the study of PPs independent of PT and even APRPP, the latter being parallel to Chomsky’s notion of true versus accidental universals (TU versus AU). In principle, it may be the case that there are a subset of PPs that occur in CS but by accident are not attested in AL; we may call this subset the unactualized or nonphonologized PPPs (NPPPs). Likewise, there may be phonological principles (PhP) or rules, as reviewed by Hyman (1976) which have independent status from PPPs (as well as APRPPs). We may call this subset of phonological rules (PRs) the non-process phonological rules of language (NPPR), and those which are true universals but not attested the non-actualized ones (NNPPR).

Summarily, this leads us to the following clearcut situation. The innate LAD provides us with PPPs and PhPs, the former operating in CS and AL, the latter in AL. Within each, there are both NPPs and NNPPRs, due to accidental aspects of the world. The PhPs will select from PPPs to create APRPPs, yet they will also create NPPRs. There is also the possibility that certain PPPs are not operated upon by the PhPs to create PRs, resulting in the subset NPPP. If so, these speculations would substantiate the Ho in (1). A careful empirical investigation would ultimately verify whether such a conclusion is justified.

Conversation as Paranoia—Hugo Aelgh
An Iñupik Linguistic Fragment (or, the Last Grammarian)—Metalleus
Lingua Pranca Contents