PHONETIC LAW (Gk. Owe-rem:6s. phone'di kos, relating to sound, from owvi), phone, sound, voice; connected with pl7nni, Lat. fori, to speak, 0110. ban, Ger. Bonn. AS. Lunn, Eng. Lan). in comparative linguistics, a formula which sums up a certain phonetic eorrespondenee or a certain number of such correspondences. The term law in this sense is, therefore, essentially different from its application in mathematics or chemistry, and approaches rather the usage of the word as it is employed in psychology. A given pinmetic law, moreover, is required to he opera tive only under given conditions in a given dialect or dialect-group during a given period. As an example of such a law we may take the represen tation of the Indo-Germanie 0 which becomes in 6, which remains unchanged in Gothic, Icelandie. Ill l Saxon, and Anglo-Saxon; as San skrit Greek opdrnp, Latin frnter, 'brother.' lint Gothic briar Icelandic broPer, old Saxon brothar, Anglo-Saxon troor. Further, it is a phonetic law in Old lligh German that such an 6 becomes no, which is changed in New nigh German to 0. so that we find correspondirw to Gothic br4par, Old High German brooder, New High German Bender. On the other hand, Indo Germanic 0 becomes a in Old Church Slavic and written al in Lithuanian. as Sanskrit bhratar, 'brother.' Old Church Slavic bratrfi. Lithuanian broils. .1.: the science of comparative linguistics developed. it was found that sound-changes fol lowed certain rules. The very earliest investiga tors, however, seem to have paid little attention to the problem whose enunciation and discussion was destined to mark an epoch in the history of the science, the question as to the invariability of phonetic law. Yet even then it was realized that etymology (q.v.) without strict adherence to phonetic law was capricious and unscientific. The existence of such law was emphasized by the famous discovery known as Grimm's law (q.v.), which was in its turn supplemented and accen tuated anew by the enunciation of Grassmann's law (q.v.), Verner's law (q.v.), and the palatal law, which postulated the existence of e in Pre Indo-Germanic. In 1876 the logical result of phonetic investigations up to that time was given by Leskien, when be stated the theorem that phonetic laws know no exceptions. This was the keynote of the neo-grammarian movement. (See PHILOLOGY.) Osthoff and Brugmann two years later stated the principle in its definite form as follows: "All sound-change, in so far as it ope rates mechanically, is carried out according to in variable laws." By this modification of Leskien's statement the necessary scope was given for analogy (q.v.) in controverting the action of the laws. Exceptions were, therefore, only apparent, not real, and they were to he explained either by the operation of analogy or as the results of laws yet undiscovered. This latter point had already been emphasized by Leskien, as Whitney had laid stress on the importance of in explana tion of many difficult linguistic phenomena. The
point at issue was unfortunately shifted, and the gulf between the old and new schools of gram marians widened almost hopelessly. The conten tion became not the nature of phonetic law, wherein all might, it would seem, be in harmony, but whether the laws were invariable or not, a problem which is yet unsolved. Without going into the details of the long discussions which fol lowed after 1S7S, where Georg Curtius and Brug mann were the protagonists for the old and new schools respectively, it may be said that the net result has been a distinct gain for the neo-gram marian movement. Phonological and etymologi cal investigation at present tacitly accepts the in variability of phonetic law. At the same time, it is not as rigidly hound by dogmatic adherence to this principle as was the case when the en thusiasm of its novelty was potent. The influence of the doctrine has, then, been most beneficial in enforcing accuracy in the application of the laws, and compelling exact and full explanation of any deviation, however slight, from their normal ac tion. On the other hand, sufficient attention has not yet been devoted to the explanation of the laws themselves. Why, for instance, to revert to the example already given, does Indo-Germanic it become 4 in certain languages, a in others, and in others still remain unchanged? Phonetic changes, instead of being absolutely invariable, are only relatively uniform and obey only general ten dencies. These tendencies are governed to a large extent by physiological requirements, and fur ther by psychological processes, especially the associative faculties of analogy, imitation, and the like. Initially the changes are individual istic. If the phonetic change in question is physiologically possible and pleasing to a speech comniunity. however it may be adopted imitatively, and may then be extended analogi and being accepted by a widening range of speakers, it may develop into a general tendency. become relatively uniform, and thus be made a phonetic law. These conclusions are confirmed by the study of living, spoken dialeets, where the artificial uniformity of script does not mislead the investigator, as is too often the case with languages which have been preserved only in literary forms.
Consult: Curtius, Zur Kritik der neuesten, sprachforsehung (Leipzig. 18S5) ; Brugmann, Zinn heutigen Stand der Sprochtcissenschaft 1SS5) ; Schuchardt, Utber die Lautgesetze 18S5) ; Delbriick, Einlcituay in dos Sprach stadium ed.. Leipzig, 1S93) ; Paul, Prinzivien der Sprachgcschichte 13d1 ed., Valle, 1S9S) ; Oertel. Lectures on the of Language (New York, 1901). See GRA SSMAN LAw : GRIMM'S LAW; PHILOLOGY; PHONETICS; VERNER'S LAW.