The marriage of phonetics and phonology

Abstract

The distinction usually made between phonetics and phonology is that phonetics studies the physical or physiological aspects of speech, including its articulatory, aerodynamic, acoustic, auditory, and perceptual aspects, whereas phonology is concerned with accounting for the variation in speech sounds in different but related languages and dialects, and within a given language in the environment of different morphemes, different positions within an utterance, word, or other speech sounds. Phonology also derives generalizations about particular languages and common patterns of speech sounds in all human languages. More modern manifestations of phonology purport to characterize the psychological or even the genetic underpinnings of language, including its sounds. Table 1 presents some examples of generalizations that might be deemed exclusively the domain of phonetics (in the first column) or phonology (in the second column). In the earliest known descriptions of speech sounds in languages there was no distinction between phonetics and phonology. Panini, for example, who worked in approximately the 5th c. BPE, compiled a magnificent and detailed description of Sanskrit speech sounds and their contextual variations [1]. The beginnings of the differentiation between phonetics and phonology, however, probably began in the 19th century and accelerated in the 20th. A number of cumulative developments account for this: . The 19th century saw the success of historical linguistics in establishing family relationships between languages by discovering systematic phonetic relationships between numerous words in two or more languages. Table 2 gives some examples. This permitted Rask [2], for example, to extract the generalizations equivalent to ‘voiced stops in Latin and Greek corresponds to a voiceless stop articulated at the same place in Germanic languages’ (such as English). . In 1861–62 Schleicher [3] posited abstract parent forms—marked with an asterisk—for cognate words in the parent language, now called Proto-Indo-European (see column 4 in Table 2). . In 1878 Ferdinand de Saussure [4] posited the existence of sounds (now known as ‘laryngeals’) in Proto-Indo-European, these sounds being known only by their effects on other sounds since they had disappeared in the languages on which Saussure based his study (Subsequently some of the posited ‘laryngeals’ were found in Hittite texts). . In the early 20th c. Saussure [5] taught that beside the immediately observable facts of a language’s sounds there was also an underlying reality, its structure or system, that needed to be established. . The concept of the phoneme arose from the teachings and practice of Baudouin de Courtenay [6], Saussure [5], and Sweet [7], the phoneme, which might have a variety of contextually-determined phonetic variants, was regarded as the minimally contrastive unit in the make-up of words and morphemes. To give one simple example: the initial sound in ‘‘pip’’ [pIp] and the second sound in ‘‘spill’’ [spIl] were said to be members of the same phoneme /p/ (identified as a phoneme by the use of the forward slashes); the first member being aspirated and the second one not. Conceptually the phoneme is parallel to the one in historical linguistics where, e.g., a single parent sound p is reconstructed for the variants found in related languages, i.e., [p] in Latin and [f] in Germanic (see above). The phoneme was conceived of as the psychological ‘parent’ form of the contextuallydetermined phonetic variants, e.g., a phoneme /h/ in English was the psychological source of the phonetic variants [h] in ‘‘how’’ [hAu] and [ç] in ‘‘hue’’ [çju]. . With the rise of generative phonology [8] variant forms of morphemes such as profane /profejn/ profanity /profænIRi/ were presumed to be derived from a common abstract underlying form /prof nIti/ in the mental lexicon. Thus it was claimed that a