Two theories are held concerning the phonetic nature of the laryngeal phonemes. Ferdinand de Saussure originally conceived of the hypothetical phonemes, later partially identified with the laryngeals of Hittite, as sonants. He regarded the long vowels of heavy bases as the result of earlier diphthongs or of full-grade vowel plus consonantal allophone of one of these sonants plus a consonant. PIE he considered to be the vocalic allophone of the sonant standing between consonants in zero grade of the root. Saussure also derived the ‘long vocalic sonants’ of Proto-Indo-European by a development parallel to that which produced the long full-grade vowels. His view that what we now call laryngeals were phonemes having both vocalic and consonantal variants was on the whole retained by Albert Cuny, Walter Couvreur, and Holger Pedersen, although the number of laryngeals and the details concerning their development varied from scholar to scholar.

The verbs bregða ‘to dart, brandish’, brá: brugðum, brugðenn, and drekka ‘to drink’, drakk: drukkum, drukkenn belong to the 3rd ablaut series. The past participles brugðenn: drukkenn are phonetically irregular, in that we should expect here the α-umlaut of the radical vowel u > o: *brugðanar > *brogðenn (cf. ONorw. brogðenn), *drunkanar > *drokkenn (cf. *sunkwanar > sokkenn ‘sunk‘). It is obvious that the radical vowel u in brugðenn: drukkenn does not represent the original u, but a later analogical restoration of u in place of *o. This restoration has never been satisfactorily explained. One is tempted to explain it as borrowed from the preterit plural stem (brugð-um: drukk-um), but this hypothesis is unsatisfactory because it leaves unexplained the reason for a borrowing restricted to these two forms. Noreen and Bethge assume the borrowing theory in the case of drukkenn, but do not apply it to brugðenn, for which they give no explanation. It is difficult to understand this inconsistency, since in the two forms the phonetic conditions are apparently parallel so far as the a-umlaut of u > o is concerned.

Recently, in a workshop of members of the Summer Institute of Linguistics in Peru, working with about twenty-one languages from nine language families of the area, I had an opportunity to observe the fruitfulness of one set of exploratory concepts postulated in Part II of my Language in relation to a unified theory of the structure of human behavior. The relevant section of Part II states: ‘Our empirical data, however, suggest that we must assume contrastive differences in abdominal pulses of types not postulated by Stetson. ... I am convinced, on kinesthetic evidence, that we must for the abdominal pulse, as for the syllable pulse, postulate types which are (a) ballistic, (b) controlled, (c) beginning ballistic and ending controlled, (d) beginning controlled and ending ballistic’ (57a).

A free allophone is that allophone of any phoneme for which no feature or articulatory movement is determined by its environment. By extension, an allophone may be said to be free with respect to any undetermined feature or phase even though some other of its features or phases may be determined.

It is partly a matter of terminology. It seems advantageous under certain circumstances, including the pedagogical, to use 'free' as the antonym of 'determined'. If one wants to press the concept of determination to such minute detail as to include all allophones, i.e. to say that no two realizations of a phoneme can be same except in identical environments, then 'determined' has no antonym. But there is ample precedent for drawing a line, perhaps best defined as that between perceptible and nonperceptible. The difference between the /w/s in a queen and chickweed is readily perceptible; the difference between the /k/s, if not so clearly hearable, is certainly perceptible in the proprioceptive sense of the observer's empathic analysis. The difference between the /w/s in a queen and between or in chickweed and shirtwaist is not, however, perceptible, nor, as a matter of labialization, that between the /k/s in a queen and a quart or in chickweed and backwards.

Thanks to the means available for speech analysis and synthesis, there is at present considerable knowledge of the acoustic differences whereby we distinguish among the various sounds of English. We know, for example, what the more important acoustic cues are by which listeners decide whether a stop sound is velar, alveolar, or bilabial. In the case of the stops, moreover, we know by and large what the acoustic differences are between the members of each of the homorganic pairs p-b, t-d and k-g (in other words, between the so-called voiced and voiceless stops). But we are as yet unable to state with any exactness the extent to which each of these acoustic differences separately affects our ability to discriminate between the two kinds of stops. Moreover, where progress has been made in this direction it is certainly in some measure due to the fact that we have dealt chiefly with the stops in those positions where the voiced-voiceless contrast is phonetically maximal, that is, in initial and final positions.