GIFT OF . 
 
'U^f, -fL. u^y^^ ^^ ' -/ f. /. y^^.,^:^:;^ 
 
<Xi)&[ 
 
 [Reprinted from the Publications of the Modern Language Association of America, 
 Vol. IX, No. 2.] 
 
 /tWA^ 
 
 THE PSYCHOLOGICAL BASIS OF PHON ETIC L AW 
 AND ANALOGY. -^ """ 
 
 "All's love, yet all's law." — Browning, Saul XVII. 
 
 The subject of Phonetic Law and Analogy belongs to 
 Psychology as well as to Philology ; it originated, however, 
 and was for a long time kept within the domain of the latter 
 discipline. When it first came into the foreground of our 
 interest, it was agitated by scholars in historical philology. 
 The experience of daily linguistic research suggested both the 
 question and the answer, and in fact the main importance of 
 the whole problem was seen in the influence which its solution 
 must have upon the methods of detailed investigation. 
 
 Do phonetic laws act with absolute consistency or not? 
 and how far may analogy be recognized as an agency in 
 language-life ? These were the points at issue. In a general 
 way it was a movement from casualty to causality. To say 
 that any one man started this movement a few years before or 
 during the decade from 1870 to 1880 would be unfair. There 
 is an uninterrupted course of development from Schleicher to 
 Brugmann, from Whitney to Paul. On the other hand, how- 
 ever, it would be equally unfair not to grant that during the 
 time mentioned a vague impression, a sub-conscious motive, 
 evolved into the clear statement and the strict observance of 
 a principle. 
 
 The fact had been recognized, that the Aryan mother-tongue 
 was a language with fully developed inflections long before it 
 split into its various idioms. Then, however, the so-called 
 Indo-European roots could be no longer considered as realities ; 
 they were henceforth grammatical abstractions and convenient 
 formulas. The mechanical dissection of individual words 
 could no longer be applied to languages which had inherited 
 these words as ready-made units, but each single word-form 
 
 1 
 
• •••-•• • 
 
 • •••••• . 
 
 • • • •« • • • 
 
 • • • • • • 
 
 
 2 G. E. KARSTEN. 
 
 must be judged as a whole in connection with and in the light 
 of related forms. It was seen — and this view was especially 
 suggested by the observation of living dialects — that gram- 
 matical systems as those of declensions or conjugations, were 
 apt to mould the shape of any individual member of the 
 system. Thus the principle of analogy found its way into 
 philology. Thereby a number of phonetic laws appeared to 
 be relieved of what so far had seemed to be exceptions, and 
 this finally led to the conclusion that phonetic laws suffer no 
 exception whatever, and that all deviating formations must 
 be due to outside influence, to analogy. The representative 
 men during this period of evolution are Schleicher, Curtius, 
 Whitney, Ascoli, Scherer, and Leskien. Each of them con- 
 tributed a large share towards the final result. Leskien 
 especially emphasized the new ideas in theory and practice. 
 In his academic lectures at Leipzig he inspired a number of 
 pupils who were soon to develop his teachings into a code 
 which must henceforth be observed in all philological work. 
 They insisted with ever-increasing energy upon the necessity 
 of adherence to laws and upon the all-pervading influence of 
 analogy. At first their statements were still of a confessedly 
 dogmatic character. They only claimed for their views the 
 value of an unproved but plausible theory ; in the absence of 
 real arguments, they appealed to the self-respect of philologists 
 and enlisted the dignity of their science in behalf of their 
 doctrine. Gradually, however, they became more and more 
 aggressive ; in the preface to their Morphologische Untersuch- 
 ungen 1878, Osthoff and Brugmann published a " confession 
 of faith " demanding adherence and scorning opposition. In 
 turn the new theory was called a fad (Modethorheit), it met 
 with ridicule, and w^orse than that, with much passive resistance. 
 But soon the " aper9u " was supported by proof. It was the 
 principle of analogy that first came to be well understood, and 
 was explained by Paul, in the fourth volume of his and Braune's 
 Beitrdge. 
 
PHONETIC LAW AND ANALOGY. 6 
 
 The consistency of phonetic laws still remained a matter of 
 personal impression. Neither the inductive nor the deductive 
 method yielded conclusive arguments. In 1878 Osthoff made 
 a somewliat premature attempt to present the subject before a 
 larger public. He gathered the inductive evidence so well 
 that almost nothing new has since been adduced ; but it all 
 pointed only at a probability. On the deductive side, Osthoff 
 utterly failed. It must even be said that in ascribing a decisive 
 influence to climatic conditions he made a step backwards. 
 That theory had been sufficiently exposed by Whitney. 
 OsthofFs fatal mistake lay in his neglect of the psychological 
 element in phonetic transitions. Misled, perhaps, by the 
 growing importance of speech-physiology — Phonetics — he 
 saw one-sidedly in phonetic laws the work of the nerve-muscle 
 apparatus only. Here again, as in the case of analogy, the 
 first true light came from Paul. In his Principien der 
 Sprachgeschichtey 1880, the whole phenomenon of speech was 
 first treated on its proper ground, namely upon the basis of 
 philological experience combined with a serious study of 
 modern psychology and an exact knowledge of the physio- 
 logical genesis of speech-sounds. Through PauFs thorough- 
 going investigations our insight into the true nature of 
 Language has been much deepened ; based, as they were, upon 
 a close observation of reality, they did away with many old 
 prejudices and misleading abstractions. As Whitney's lin- 
 guistic publications are of a distinctly different character, it 
 is no injustice to our sound and sober Yale philologist to say, 
 that PauPs work was epochmaking in regard to the study of 
 every aspect of language life which it touched. Not that in 
 it all riddles were solved ; but every problem was defined, 
 reduced to its real nature and prepared for solution, if not 
 solved. A solid ground was now gained for further discussion ; 
 the dogmatic tone ceased on both sides, and soon a number of 
 fruitful essays appeared, throwing more and more light upon 
 the various sides of our problem. The men who took a 
 prominent part in the discussion are Delbriick, Tobler, 
 
4 G. E. KAESTEN. 
 
 Neumann, Misteli, Jespersen, Brugmann, Grober, Wundt, 
 Schucharclt, Kauffmann. In America some phases of our 
 subject have been treated by Bloomfield, Easton, Goebel, 
 Tarbell, Wheeler and the author. In the meantime, however^ 
 one important piece of inductive evidence had also been 
 obtained in favor of the law-theory — an argument which Ost- 
 hoff was hardly in a position to quote — that is the enormous 
 gain which has come to philology through the new dogma 
 as applied by the "neogrammarians^^ themselves and by those 
 who sooner or later adopted the same method in practice, though 
 not always in theory. Not only has the philological work of 
 Osthoff and Brugmann, of Braune, Paul, Sievers, of Bechtel, 
 Collitz, Schmidt, Kluge, Noreen, Neumann, Meyer-Liibke, 
 Schuchardt,^and others developed exceedingly clear and lasting 
 results, but theirs has practically become the working method 
 of all philologists ; nobody would to-day simply register an 
 exception to a rule without making an attempt to account for 
 the same. Some philologists even seem to have already forgotten 
 that there ever was any question about the consistency of pho- 
 netic laws. In fact the problem is now more especially a psy- 
 chological one ; it has largely lost its methodological interest, 
 while the solution itself is still a matter of controversy and 
 while a few may even go so far as to deny both the truth and 
 the novelty of the law-theory. Its most successful opponent 
 is undoubtedly Schuchardt. In an admirable pamphlet and 
 in several articles in the Litei^aturhlatt he has indeed shown 
 that some of the fundamental assumptions in PauPs argumen- 
 tation are untenable. 
 
 In now discussing the subject itself we think it advisable 
 to present our own views and observations somewhat explicitly 
 rather than to offer a disconnected criticism of individual 
 arguments. In this way we shall have to repeat many things 
 that have already been said by others, especially by our revered 
 
 ^ This list is not meant to be exhaustive or representative ; it suggested 
 itself to the author at random, merely on the ground of his own studies. 
 
PHONETIC LAW AND ANALOGY. O 
 
 teacher Paul, but, in turn, the points in which we differ from 
 our predecessors will thus stand out in their true light and, 
 moreover, we hope to give such a presentation of our subject 
 as will be readily understood by those who may not, as yet, 
 have paid attention to this important question. First of all, 
 let us see what are the undisputed facts which historical phi- 
 lology teaches us concerning the effects of phonetic changes 
 and of Analogy upon the spoken language-material. 
 
 Strictly phonetic changes, that is, changes which seem to 
 affect the sounds of a language regardless of the meaning of 
 the words in which they occur, usually cause a slow graduation 
 of one sound or group into another, and the change of a sound 
 in any one word is, as a rule, accompanied by a parallel change 
 in a number of other words. A closer examination usually 
 enables us to ascertain the circumference of such a change by 
 stating the phonetic conditions under which the transition took 
 place. To be sure we can never with certainty predict the 
 change in any one word. We are strictly limited to a posteriori 
 statements, and if we call our abstractions a rule or law, we 
 must not forget that the speech -material will fit into our rule 
 only in so far as our rule has been made to fit the speech- 
 material. The fact, nevertheless, remains of importance, that 
 phonetic changes usually affect a sound in whole groups of 
 words. 
 
 Analogy, as the word is most commonly used in philology, 
 seems to operate primarily with larger speech-units, with words 
 and word-elements not only as phonetic productions, but also as 
 expressive of their respective meanings. Through the medium 
 of the word-element analogy influences the sounds of the word ; 
 it causes not a gradual transition of sounds, but it may bring 
 about a substitution of word-elements, and its work in any one 
 case does not immediately seem to involve a change of any 
 other word. — It now becomes our object to find out, how and 
 why both these factors come to manifest themselves in Lan- 
 guage, and in order to answer this question, we should realize 
 exactly what Language indeed is. We must study the result. 
 
6 G. E. KAESTEN. 
 
 the Spoken word in Phonetics and historical philology, and 
 the subject, speaking man, his physiology and psychology, in 
 order to understand the act of speech. 
 
 Human language is our agency of expression and of inter- 
 communication by means of sounds that are produced by the 
 vocal organs. Through various motions of our glottis, our 
 palate, tongue and lips, the air from the lungs is forced into 
 different vibrations which strike the ear of the hearer as sounds. 
 A word then, in so far as it consists of sounds, is merely a series 
 of vibrations which touch our ear for a moment, then eradiate 
 and, as sounds, die away ; it is the transient result of an action, 
 it has no duration, yet it has its causes in the past and its 
 consequences in the future. That same word has been pro- 
 nounced before ; we may recognize it ; it was intended by the 
 speaker to indicate something to us, and it does so ; we may 
 reproduce it on our part for the same purpose. What then is 
 going on in us that enables us to recognize a word and to 
 understand its meaning, to recall it again, and to reproduce it ? 
 Reviewing the elements of psychology we remember that the 
 sensory nervous system with its centre in the brain and with 
 its functions differentiated into several more or less distinct 
 branches is the channel through which man receives and real- 
 izes the influx of impressions from the world around him. All 
 these branches with their innumerable fibres are constantly at 
 work and simultaneously conduct their impressions to the cen- 
 tral organ, like so many reporters who are constantly wiring 
 from various places at the same time their experiences to the 
 central oflice. These reports are recorded and preserved ; ^ they 
 may and they do drop into unconsciousness ; the vast majority 
 of them never become conscious at all ; yet they all have their 
 influence upon the individual ; favorable circumstances may 
 call or recall them into consciousness or into a state of more 
 
 ^ Little is known, as yet, about the real nature of this process, and if 
 in the following lines we use expressions like "impress, imprint, deep" etc., 
 we are merely using figures of speech which are not meant to contain any 
 intimation as to the way in which our mental activity is carried on. 
 
PHONETIC LAW AND ANALOGY. 7 
 
 or less vivid and active subconsciousness. The latter is suffi- 
 cient for the purposes of speech, it is indeed the very province 
 within which the function of speech lies. The central point, 
 the " Blickpunkt '^ of consciousness is occupied by the idea to 
 be expressed, while the act of speaking is performed on the 
 very edge of the wider sphere, the "Blickfeld^' of conscious- 
 ness, and partly, so far as the mechanical execution is con- 
 cerned, within the realm of unconsciousness.^ 
 
 Now the readiness with which a sensation or perception may 
 come into consciousness, or into what we may call cooperative 
 subconsciousness, is determined by various circumstances : by 
 the condition of our mind, its momentary susceptibility as well 
 as the direction of its main attention, and by the nature of 
 the impression itself, its absolute energy as well as the number 
 of times the impress is repeated. Upon all these factors depends 
 the facility with which the mechanical retention of impressions, 
 the general faculty of memory — " Gedachtniss " — can lead to 
 the recalling — ^' Erinnerung " — of individual impressions. 
 Among these arguments there is one which calls for further 
 explanation, as it is of extreme importance in speech, and at 
 first sight not self-evident. Namely, how can repetition take 
 place in such a way that the new impression fits into, and 
 renews or strengthens the memory picture of a previous 
 impression, when in reality, considering the incessant change 
 of subject and object, no two impressions can ever be perfectly 
 equal ? They can indeed only be more or less similar to each 
 other, but they are nevertheless connected in our mind, and for 
 this reason : similarity means, of course, partial equality, 
 equality of parts, and the apperception of equal impressions 
 into the same memory-picture naturally involves the merging 
 into one another of the corresponding equal parts of two 
 similar impressions. Of two perceptions, the one may cover 
 the province of A -|- B, the other that of A + C ; then the A 
 
 ^ This does not preclude the possibility of conscious interference with the 
 development of words. That factor may have to be considered in a study 
 of results, but the usual psychological phenomena are not thereby altered. 
 
8 G, E. KARSTEN. 
 
 of the second will be apperceived into the memory picture of 
 A of the first impression ; united A being imprinted twice, 
 may thereby stand out more vividly even, than either B or C. 
 At the same time united A will act as connecting link between 
 A + B and A + C. The phenomenon of apperception thus 
 immediately leads to that of association, and it is obvious that 
 an association will be more or less intimate in proportion as 
 the equal connecting or the unequal connected parts prevail. 
 Practically, now, extreme similarity has the same effect upon 
 our mind as perfect equality : the impressions cover each other 
 completely and the composite result is simply a strengthening 
 of the old memory-picture. This is due to two circumstances : 
 namely, subtle as our nerve-brain apparatus may be, it is not 
 absolutely perfect. There is a threshold of sensitiveness ; 
 minute impressions are not recorded; they are lost in the 
 apparatus. Hence the element of personal error in all per- 
 ceptions. This comparative dullness of our nervous system, 
 of course, accounts only for the mere possibility of similar 
 impressions being recorded as equal, while it would not preclude 
 the opposite occurrence, namely, the deviating of two percep- 
 tions from each other beyond the actual difference. How- 
 ever, as a very important fact in this connection, it has been 
 recently ascertained,^ that in every new perception the element 
 of personal error tends to work in the direction towards a 
 previous, accustomed memory-picture. Through the habit of 
 apperception and association we seem to have acquired a 
 tendency to emphasize equalities, to neglect differences, to 
 allow an old memory-picture to immediately over-shadow a 
 new perception, so that we do not realize the actual difference 
 between the old and the new. It is this fact, then, which 
 causes extreme similarity to appear as equality, the minute 
 differences remaining unrecorded. Of course, as the actual 
 discrepancy between similar impressions increases, it will make 
 itself felt more and more distinctly in the composite impress. 
 
 ^ Cf. Leuba in Clark University Minor Studies in Psychology, I. 
 
PHONETIC LAW AND ANALOGY. 9 
 
 The latter will then unconsciously, or more or less consciously 
 deviate from the previous imprint. As long as the element of 
 equality by far prevails, the congruous parts, having the 
 advantage of repetition, will alone constitute our clear concep- 
 tion, while the deviating elements largely remain unconscious, 
 or form about our conceptions a circle of vague suggestiveness. 
 On the other hand, impressions which have only small and 
 insignificant parts in common and which are essentially different 
 from each other will overlap but little, being in the main 
 registered separately. Such perceptions may be so loosely 
 associated with each other that they need especially favorable 
 circumstances for calling each other up at all. Here belongs 
 the peculiarly complicated yet loose connection which exists 
 between the simultaneous impressions of the different senses. 
 They have in common the element of time, that means, of 
 surrounding, the general condition (" Gemeingefiihl '') of the 
 individual. They are indeed altogether a physiological unit, 
 because the various sensory centres are, of course, directly 
 and indirectly connected among each other in the indi- 
 vidual.^ 
 
 Yet the unit is — as in the case of most perceptions and 
 sensations — distinctly a complex one, and experience shows 
 that the different constituents, as they are introduced by the 
 various channels of the nervous system, only connect very 
 loosely with each other. A sight-picture does not normally 
 impart any distinct impression to the acoustic centre. Indeed 
 
 ^ In fact from a physiological as well as a historical point of view, there is 
 in man a double connection between the seats of the various senses : the one, 
 of a higher order, located mainly in the cortex in which all the centripetal 
 and centrifugal nerves are combined into a more or less self-conscious 
 unit; this is best developed and capable of highest perfection in man, 
 diminishing and graduating away in the lower animals ; — the other, located 
 in the lower parts of the brain, the cerebellum and the medulla, the 
 province of reflex actions between the diflferent branches of the system ; this 
 interrelation is stronger and more depended upon in the lower animals than 
 in the more developed ones ; it is evidently the remnant of the original 
 uniformity of the whole nervous system. 
 
10 G. E. KARSTEN. 
 
 the main effect of one sensory centre upon the others seems ^ 
 to consist in a general agitation through the whole system, 
 arousing or dulling it into a certain state of susceptibility. 
 Thence it comes that when one of the parts of such a simul- 
 taneous impression is in the future repeated, it does not neces- 
 sarily call up again the whole previous group. The connecting 
 element is not strong enough to link the different parts imme- 
 diately together; yet the oftener the latter are introduced 
 together, the prompter and the more unfailingly will the con- 
 nection work — it is strengthened by repetition. — This obser- 
 vation finds its reflex in the physiological fact, that when 
 through a lesion of the brain the usual path between two parts 
 is destroyed, and a new connection is formed, the new line will 
 at first conduct only imperfectly, but through habit the 
 transmission becomes more and more correct and prompt. 
 With these premises we shall readily understand two phases 
 of language-learning, that of remembering and of under- 
 standing words. — The child hears the word mama while its eyes 
 reflect the features of the mother and while through her care 
 it is being made comfortable and happy. That means, at the 
 central organ a group of simultaneous reports are received : 
 there is the sound-picture of the word mama, there is the sight 
 picture of the mother's features, and there are, through various 
 other channels, reports of pleasant sensations. The sound- 
 picture mama is thus connected with the remainder of the 
 group. Now the situation passes by, the associated impressions 
 
 ^ This question, like many others relating to the working of the nervous 
 system and the brain, is yet far from being settled. Numerous cases are' on 
 record of a very distinct and immediate inter-relation between the different 
 senses. The author can here speak from his own experience ; the vowel 
 sounds most vividly suggest to him colors and vice versa : a = red, o = 
 blue, e = green, a =: yellow, i = white, u = purple-black, etc. Our languages 
 abound in precipitates of such and similar connections : loud colors, soft 
 sounds, etc., etc. — But however immediate, unavoidable, and natural, such 
 mutual interrelations may appear to us, it seems difficult to prove, that they 
 are not the result of early associations brought about by personal experi- 
 ence of the individual or of his ancestors. — Yet see now Krohn, " Pseudo- 
 Chromesthesia," Am. Jour, of Psychology, V, pp. 20-42. 
 
PHONETIC LAW AND ANALOGY. 11 
 
 drop into unconsciousness and there they remain, as memory 
 pictures. The following times, whenever the word mama is 
 heard, the situation again includes the shape of the mother 
 and, besides, usually a number of pleasurable sensations. The 
 latter, though mostly pleasant will, however, vary greatly in 
 detail, while the sight-picture of the mother and the sound- 
 picture of mama mostly form a part of the group. These two 
 most stable elements form thus the most distinct impressions 
 and are preeminently connected with each other : the word 
 mama finally means nothing else distinctly but the mother, 
 while all the other impressions oscillate about the sight- and 
 sound-pictures, uniting what they have in common, and thus 
 surround the word mama or mother with that halo of sacred 
 sweetness which it possesses everywhere. In this way the 
 sound-picture of every word is fixed in our memory, receives 
 its distinct meaning and also, to a large extent at least, its 
 collateral associations, its suggestiveness and general character, 
 or what might be called its topographical place in the indi- 
 viduaFs vocabulary. This process is fundamentally the same 
 everywhere : in the child, when it learns its first words from 
 parents and servants, in the youth, when taught to accelerate 
 the acquisition of new material by paying special attention to 
 his language, in the foreigner who tries to learn something from 
 everybody, in all men as they are gaining a fuller and more 
 correct use of their mother-tongue. There is no reason to 
 believe that the process was essentially different at the time 
 when human language was in its first formative period. 
 
 Our next question is, how do we learn to pronounce and to 
 use words correctly ? Bearing in mind that we do not speak 
 here of the first origin of speech, but of the normal nature of 
 an already existing language, we will first recall the fact, 
 emphasized by Whitney, that we must hear a word from others, 
 before we can think of pronouncing it ourselves ; hearing 
 precedes and elicits speaking; the first foundation of our 
 vocabulary consists of words whose sound-pictures we have 
 already developed in our memory. We reproduce these words 
 
12 G. E. KARSTEN. 
 
 as we heard them ; we imitate. The kind of imitative repro- 
 duction however, is of a somewhat complex nature. We have 
 only the final aim, the result to guide us, while we know next 
 to nothing about the way in which this result was and can 
 again be obtained. We cannot see the motions of the vocal 
 organs in other persons clearly enough to be guided by them ; 
 nor have we control enough over our own organs, consciously 
 to repeat the motions of others, even if we could see them. 
 How, then, does the child contrive to pronounce any given word 
 for the first time ? Indeed another question would first arise. 
 What is it that induces the child to use its vocal apparatus 
 at all ? We can only state the fact that children can and do 
 cry almost before they can perform any other motion. That 
 means, they inherit a predisposition to produce sounds with 
 their vocal organs ; the motory nerves which control the latter 
 respond easily to every sensation that is reported at the central 
 organ. Therefore whatever agitation may be felt anywhere 
 in the nervous system, it is especially the vocal apparatus that 
 is set in motion, to restore the equilibrium. Moreover, there 
 seems to be an especially close correspondence between the 
 motory nerves of the vocal organs and the sensory acoustic 
 centres ; the former are called upon to react, as soon as an idea 
 agitates the child's mind and in connection with it the sound 
 picture of a word arises. As the eye governs the hand, thus 
 the ear controls the vocal apparatus. The tendency, at least, 
 to develop this interrelation is also inherited and thus far may 
 it be said that Language is natural to man. 
 
 But what enables the child to move its vocal organs in the 
 right way so as to reproduce a certain sound ? This ability 
 must be acquired by each individual through practice, just as 
 the execution of almost every motion must be learned. Yet 
 the child undergoes a good deal of preparation for its first task. 
 Long before it tries to reproduce any one word, it has been 
 uttering many different sounds and sound-groups at hazard, 
 without any purpose but that of motion as an outlet of agi- 
 tation. Of course, it has at the same time been hearing the 
 
PHONETIC LAW AND ANALOGY. 13 
 
 phonetic effects of those motions, its own cries and sounds. 
 An association, an unconscious grouping has taken place 
 between the various sounds and the sensation of the muscular 
 motions that produced the former. The child says many times 
 may ma, ma, papapa before it notices that mama, papa have 
 any meaning. These first sounds are caught up by the parents, 
 they are repeated and interpreted as mama, papa — in fact, 
 these words, most likely, owe their origin to baby-talk. Now 
 the child hears his own mamama repeated as mamay he gradu- 
 ally — as we saw above — connects the word with its meaning 
 and thus it is that when he first wishes to say mama, the way is 
 prepared, the vocal organs move about in the right way, because 
 the sensation of the required motion is already developed. 
 The scope of phonetic acquisitions is thus gradually enlarged, 
 until all the sounds of the mother tongue are at his disposal. 
 To be sure, none of the first reproductions of a word are 
 entirely successful ; many attempts may be necessary, before 
 the resulting sound is in harmony with the sound picture in 
 the speaker's mind. Little by little, however, the motion and 
 the sensation of motion adapt themselves and become so 
 familiar, the connection between the idea, the sound-picture, 
 the sensation and the execution of the motion, becomes so per- 
 fect, that the conception is expressed by the pronounced word 
 with a promptness which approaches the immediateness of a 
 reflex motion.^ 
 
 There are other factors of importance for the development of 
 speech, we may mention especially the motive of sym})athy on 
 the part of the hearer;^ but what has been said, will be 
 sufficient — as it was necessary — to help us to understand our 
 
 ^ In this connection we mention the controversy, if such it may be called, 
 between Max Miiller and Whitney. Max Miiller has published and re- 
 published large volumes in which he assumes and tries to prove that Lan- 
 guage and Thought are identical ; Whitney has several times taken occasion 
 to expose the absurdity of this assertion, but apparently without any effect 
 upon Miiller whose recent publications still essentially repeat the funda- 
 mental errors of the first editions. 
 
 '■'Cp. Wegener, Grundfragen, pp. 13 and 66. 
 
14 G. E. KARSTEN. 
 
 special subject. It appears that it is not the spoken word that 
 changes, but the acoustic picture in our mind through its 
 transmission from one individual to the other. Treating now 
 first the especially so-called phonetic changes, it is clear that 
 the very act of speaking involves the possibility and the 
 necessity of change. We saw how Language must be gradually 
 acquired and, of course, the process of learning which goes on 
 throughout life suggests all imaginable phases of development, 
 from the first uncertain attempts of the child to the high 
 degree of perfection which the trained and refined man may 
 obtain. When we further consider that men are not all equally 
 gifted and have not all equal opportunities, we at once under- 
 stand that there must be an infinite number of speech varieties 
 in existence all the time. Accurately speaking, there are as 
 many different varieties as there are persons speaking and even 
 the language of every individual may be subdivided into 
 different kinds of speech according to the mental condition of 
 the speaker and the social atmosphere in which he is moving 
 at the time. All these speech varieties are constantly influ- 
 encing each other. Even a pronunciation noticeably different 
 from our own and one which we subconsciously feel to be less 
 perfect will affect our own speech as long as the sound pictures 
 heard promptly call up and join the memory-pictures in our 
 mind, and the result must be a new, composite sound-picture, 
 a compromise between the old and the new. However, the 
 act of speaking causes a steady change of speech in still 
 another way. While the mutual influence of the different 
 speech-varieties upon each other might be expected to finally 
 lead to some agreement, a source of incessant transitions lies 
 in the fact, that no one, not even the best trained speaker, 
 reaches absolute perfection, so as to make his own pronunciation 
 invariably and entirely harmonize with the sound-picture in 
 his own mind, Nor is the hearer always in a position to 
 receive the spoken sounds exactly as they were uttered. Both 
 our sense of hearing and especially the nerve-muscle apparatus 
 of our vocal organs are unfit for absolute accuracy ; on its 
 
PHONETIC LAW AND ANALOGY. 15 
 
 complicated way from mind to mind, the speaker's sound- 
 picture is liable to be reproduced in a somewhat altered form 
 in the hearer's mind, there modifying again the memory-picture 
 and even its own prototype, for, of course, the speaker himself 
 is at the same time hearer also. 
 
 If then a slow but incessant change is possible and una- 
 voidable, the question arises in which direction does the change 
 take place and which are the directive factors that determine 
 its course? Historical philology shows the fact that gradual 
 transitions of sounds take place in all possible directions, that 
 no general principle can be stated which would enable us to 
 predict what any individual sound or sound-group will become 
 in the future. We can only register historical facts and from 
 these we may try to draw our conclusions as to their causes. 
 Several reasons have been justly quoted which may account 
 for the direction of sound-change in a general way. On the 
 one hand the element of euphony, vague and subjective as it 
 may be, must yet be recognized as an agent that will influence 
 the production of sounds ; of course we are speaking of the 
 unconscious preference for some shade of sound that may be 
 more sympathetic than another. On the other hand, and 
 equally unconscious, the motive of ease, of economy of energy, 
 is at work shaping the pronunciation of our words. But 
 while both these factors, the acoustic element of euphony and 
 the genetic motive of ease, may underlie all phonetic change 
 in general, they yield, for that very reason, quite unsatisfactory 
 explanations of any one individual transition. The question 
 still remains, why is it that these same causes lead to a certain 
 result in one case, while at another time they produce just the 
 opposite effect ? Some scholars ^ ascribe a determining influ- 
 
 ' Scherer, Osthoff, Kauffmann. Very different from their theories is the 
 fact which Sievers first discovered, that the ordinary position of the organs 
 of speech — " Indifferenzlage " — and therefore the basis of articulation is a 
 different one in the speakers of different languages. How far this is due 
 to racial differences or to the influence of the languages themselves, remains 
 to be decided. My own observation leads me to believe that it is not only 
 the cause but mainly the result of speech -peculiarities. 
 
16 G. E. KARSTEN. 
 
 ence to descent, race differences, or to climatic conditions, or 
 to historical events ; and undoubtedly it must be granted that 
 everything that influences speaking man physiologically or 
 psychologically must modify his speech in a certain way. 
 However, it must also be confessed that as yet nothing is 
 known in general or in detail about the direction in which 
 these factors may tend to affect language. Much has been said 
 about the soft and musical sounds of the South as compared 
 with the harsh, rough Northern type, but, as Whitney has 
 rightly stated, no concrete illustrations can be quoted that could 
 not be counterbalanced by as many examples which would 
 seem to prove the opposite. Important as may be the effects of 
 these agencies, they seem to be hopelessly covered up by the 
 cross influence of other factors which lead to more immediate 
 results. They are on the one hand, it may well be j^resumed, 
 largely interfered with by the personal qualities of the indi- 
 vidual speaker, by his private condition of life, his habits, his 
 general character. Probably the poor people, the laboring 
 classes all over the globe, have more features in common that 
 are liable to modify their speech, than all the members of any 
 one nation or race, and so have the people who lead sedentary 
 lives, and so the old, and the young, the men, and the women. 
 This consideration shows how difficult it will forever be, to 
 distinguish any traces of racial, historical, climatic influences 
 within this diversified mass of individual cross influences. 
 The difficulty seems to grow into an impossibility, when we 
 remember that a safe standard for comparison is not at our 
 disposal, because no two languages or dialects possess the same 
 speech-material. Moreover, and this is of very great im- 
 portance, the influences of all the agencies mentioned are 
 superseded by that of speech-intercourse. It is pre-eminently 
 the speech-community which moulds the individuaPs language. 
 But first it may be well to call to mind what a speech-com- 
 munity is. Accurately speaking, the individuaPs speech-com- 
 munity comprises all those and only those people with whom 
 the individual actually associates, whose language he hears. 
 
PHONETIC LAW AND ANALOGY. 17 
 
 And it is clear that this means something different in the case 
 of every person. On the other hand, it is evident that all 
 these speech-circles are more or less closely interwoven with 
 each other and act upon the individual directly or indirectly. 
 This all-controlling influence manifests itself in a double way : 
 persons will speak the same language in proportion as they 
 grow up and live with each other in an evenly intimate inter- 
 course. If the Negro, the Irishman, Frenchman, German in 
 America, the Jew in Germany, is promptly recognized by his 
 peculiar brogue, it is mainly because he associates more with 
 people of his own kind proportionately than with the public at 
 large. On the other hand a community entirely homogeneous, 
 and of course, subject to the same climatic conditions, will yet 
 develop within itself several speech-varieties in proportion as 
 social groupings exist whose members associate more intimately 
 with each other than with persons of other groups; and it 
 is not in such cases the general nature of the speaker that 
 causes him to belong to the one or the other language-group, 
 but as he changes over from one social circle into another, he 
 will unconsciously adopt the speech of the new circle ; in fact 
 the same person may acquire the peculiarities of several lan- 
 guage-circles, that means, he may speak several similar, yet 
 noticeably differing languages and use them in accordance with 
 the suggestions of the momentary situation. Such facts prove 
 in a general way the predominating power of speech-inter- 
 course. Yet it may be questioned just how far and why the 
 latter succeeds in overruling individual speech-tendencies. The 
 sound-pictures in the tnemory of every individual are of course 
 nothing but the results, the composite-images of all the sounds 
 which the individual has ever heard, with a strong predomi- 
 nance of those heard last. So the language of the whole 
 speech-community is directly or indirectly the parent and 
 prototype of the individual's language. But we must remember 
 that the individual speaker himself is not only a member of 
 that very community, but on the average every single person 
 hears his own language as much as that of all the other people 
 
18 G. E. KARSTEN. 
 
 put together, and every single pronunciation of a word modi- 
 fies not only the sound-picture in the speaker^s own mind, but 
 also his sensations of motion. These two facts, the double 
 reaction of the speaker's own language upon itself and its 
 numerical preponderance over any other speech-variety would 
 strongly favor the development of individual peculiarities. 
 Yet this tendency is very successfully checked through other 
 agencies which are rooted in the very nature of human language. 
 The latter is distinctly a means not only of expression but, and 
 much more, of intercommunication. It develops out of social 
 necessities and instincts, of a certain sympathy between hearer 
 and speaker. The willingness on the part of the former to 
 listen, to understand, and to react, offers to the latter the main 
 encouragement to speak, as Wegener has well expounded in 
 his book mentioned above. And similarly the act of speaking 
 presupposes a desire to make the listener understand. Out of 
 this mutual relation grows the tendency to adapt one's own lan- 
 guage to that of the speech-communityj so that not only some 
 crude expression but as perfect a mutual understanding as 
 possible may be obtained, and all hesitancy or partial mis- 
 understanding, all loss of time or energy may be avoided. 
 This is the uniting and at the same time the main conservative 
 factor in Speech, for the result is, of course, not only an 
 increased attention and readiness to apperceive the sound- 
 pictures of others, but also a tendency, in one's own pronunci- 
 ation, to reproduce as exactly as possible the sound-pictures 
 thus received ; to be sure the succass is only relatively complete ; 
 no two individuals have absolutely the sdme pronunciation and 
 no language is absolutely stable. While in this struggle of 
 uniting and conservative against differentiating and modifying 
 elements it is impossible in a single case to determine the work 
 of either of them, some general conclusions may nevertheless 
 be drawn from what has been said so far : when every one 
 unconsciously tends to hear and pronounce as well as he can, 
 the personal deviation will be reduced to as little as possible. 
 What remains is the unavoidable result of the individual's 
 
PHONETIC LAW AND ANALOGY. 19 
 
 physiological and psychological short-comings. Now the right 
 sensation of motion will naturally be most distinctly acquired 
 and the sound-picture most exactly perceived by the most 
 refined, the most social, and the best-trained speaker. The 
 latter, however, the sound-picture, will also depend upon the 
 character of the community : when the parent prototype, i. e. 
 all the various sounds heard, do not differ much from each 
 other, then the composite sight-picture in the individual will 
 be very distinct and precise. In proportion as the parent 
 pictures were heterogeneous, the composite result will be vague, 
 and admit and be suggestive of different shades and varieties. 
 This will in turn leave the innervation without strict direction 
 and allovv more deviation. It is evident, then, that in an 
 exclusive, homogeneous, highly developed and very social 
 community the standard of pronunciation will be strict, un- 
 compromising, and the same will be longer preserved; in 
 proportion as the community extends, as it is heterogeneous, 
 as coarser elements prevail and speech intercourse is less inti- 
 mate, language will become more diversified, and it will be 
 also more susceptible of change. In reality, the net-work of 
 speech intercourse among men is very unevenly woven ; all 
 kinds of influences, social, financial, religious, political, national, 
 geographical, are apt to divide people into more or less ex- 
 clusive groups. In each of these a different sum total of 
 individual tendencies prevails, and manifests itself in different 
 shades of speech ; as the uniting influence of mutual intercourse 
 decreases, these shades between whole groups may accumulate 
 and result in the formation of distinctly different dialects. 
 We now return to the more special study of Language in 
 the individual and to the question — which are the factors 
 that determine the direction of sound-change ? • We saw that 
 from the point of view of the subject, the speaker, nothing 
 definite can be said ; it is impossible as yet, to discover any 
 trace of climatic, racial, historical influences. Somewhat better 
 results are obtained when the object, the speech-material itself 
 is considered. Upon the basis of the physiological genesis of 
 
20 G. E. KARSTEN. 
 
 sounds a classification of sound changes can be made and this 
 fact would at once suggest that there must indeed be a causal 
 connection between the conditions of the sounds and their 
 changes. There is, first of all, the most distinct and universal 
 influence which emphasis exercises upon the fate of speech- 
 sounds. Accented syllables develop everywhere in a way 
 different from unaccented ones, the latter being more subject 
 to assimilations, or to shortening. Within the various lan- 
 guages historical philology further reveals the fact — which 
 was already stated above — that sound-change usually affects 
 not individual words, but modifies whole groups of words 
 in one and the same direction, and generally the province of 
 a sound-change can be defined by describing the physiological 
 nature and conditions of the equally modified sounds. Some- 
 times, Phonetics will even, together with the process of the 
 change, disclose also its reasons : the organs of speech, subject 
 to general laws of gravity and perseverance, or owing to 
 psychological impulses which resulted from the situation, had a 
 tendency to deviate from their former lines of motion in a 
 certain direction and thus effectuated the sound-change under 
 observation. 
 
 It is in these cases a most natural thing that such a devi- 
 ation should have occurred not only in one individual but in 
 many persons at the same time, and that the change should 
 have spread and manifested itself in the language of the 
 community. In many other cases, however, the conditions of 
 the sounds that are comprised in one parallel change are not at 
 all equal or similar among each other ; sounds will undergo 
 the same change in words in which they occur under noticeably 
 different physiological circumstances, only one of which, and 
 sometimes none, can well be held mechanically responsible for 
 the change. For instance, vowels will change at the beginning 
 or end into a more palatal or velar shade before and after all 
 kinds of consonants, when indeed the on and off glide must 
 be expected to depend upon the nature of the neighboring 
 sounds. What can it be that holds these variegated cases 
 
PHONETIC LAW AND ANALOGY. 21 
 
 together and forces them all into parallel changes ? Paul says, 
 it is the one innervation, the one sensation of motion which 
 regulates the pronunciation of the sound-element concerned in 
 all the words in which it occurs ; whenever through a deviation 
 in any one word the innervation is modified, the pronunciation 
 of the same sound in all other words is regulated by that 
 modified innervation. This explanation seems indeed quite 
 satisfactory in the case of all those changes which affect a 
 sound under substantially equal conditions, so that the un- 
 conscious articulation in all words concerned may well be 
 supposed to respond to one and the same innervation. This, 
 however, is far from being the case everywhere, and the argu- 
 ment does not hold good, when the physiological genesis of the 
 sounds is so different, that not only scientific phonetics, but 
 even untrained, everyday observation shows a marked multi- 
 plicity of articulations. Moreover, while it must be granted 
 that the sensation of motion, especially when fairly developed, 
 is indeed an important unifying factor, it is at the same time 
 apparent that the superior regulative power lies not in the 
 innervation but in the sound-picture. The reproduction of 
 the sound-picture is the sole aim of the innervation ; and a 
 sensation of motion is entirely free to vary, only in so far as 
 it does not tend to alter the sound-picture but to reproduce it 
 more accurately ; when it oversteps the limits allowed by the 
 sound-picture, the overwhelming power of the speech-com- 
 munity exercises, through the medium of the sound-picture, 
 a checking, conservative influence, and what little change a 
 deviating innervation may succeed in causing in the sound- 
 picture, will henceforth determine the development of all the 
 several innervations whose results it comprises. Namely, it is 
 a fact — better known at present than it was to Paul — that 
 different articulations may lead to the same acoustic result, 
 that the various organs of speech may vicariously act for each 
 other and yet produce practically the same sound. It is also 
 a psychological fact that our sense of hearing is particularly 
 dull ; the average man is pre-eminently eye-minded. Yet even 
 
22 G. E. KARSTEN. 
 
 with the keener sense of sight we are unable to notice the 
 gradual daily changes of our friends as they grow into man- 
 hood or old age. It is not astonishing, then, that with the 
 more obtuse sense of hearing we should not perceive the 
 difference between the results of various, similar articulations ; 
 that means, that one and the same sound-picture in our mind 
 may allow its orders to be carried out by various innervations 
 in accordance with the different surroundings. All this leads 
 us to look for the main uniting element of all sound-change 
 not in the sensation of motion, but in the sound-picture. Of 
 course, as the result of an especially deviating, individual 
 pronunciation, any innervation, and also any sound-picture, 
 may split in two, or rather the deviation may start a distinctly 
 different impression. Under favorable conditions this division 
 may be concurred in by the speech-community, in every sense 
 of that word; and then we state, in the historical study of 
 results, the effects of two phonetic laws governing, each within 
 its own physiological and geographical limits, the same group 
 of sound varieties which so far had responded all to one single 
 law. Similarly two different innervations or sound-pictures 
 may become more and more alike and finally merge into 
 each other. Thus Old Latin e and I unite in later Latin in a 
 medium sound, close e, short or long, according to its position 
 in the syllable ; then, however, this e differentiates again, and 
 for instance, in open syllables becomes ei: oi, or ei: e, or ^, under 
 various conditions. In A. S. both a before nasal, and o coin- 
 cide in o; then this o splits into o and ce: e. Aryan oi and ai 
 both appear as ai in Germanic, and this again splits in some 
 Germanic dialects ; in O. H. G. it is e before r, h, w and final, 
 otherwise d; in Old Norse it is differentiated largely after the 
 same fashion. The phonology of all languages abounds in 
 examples. 
 
 We see, then, that the impulse to change lies in the indi- 
 vidual pronunciation and, through it, in the innervation, but 
 we also understand why it is that phoneticians are so often 
 unable to give a plausible physiological reason for a certain 
 
PHONETIC LAW AND ANALOGY. 23 
 
 sound-change, and why with the best of phonetics we shall 
 never be able to predict any sound-change. It is not the 
 phonetic conditions of a sound in any one word or position 
 that direct the transition ; the innervation is not allowed to 
 change in accordance with its own physiological motives 
 exclusively ; but the individual innervation is, together with 
 possibly several other sensations of motion, checked and con- 
 trolled by one sound-picture and in the final change of the 
 sound-picture we have either the result of the composite 
 tendencies of all the innervations concerned, or else, at best, 
 the outgrowth of some one tendency which for some reason 
 had a predominating influence on the whole group. To this 
 explanation the same objection might be offered which has 
 been raised against PauPs theory ; namely, it has been doubted 
 whether indeed one and the same innervation regulates the 
 pronunciation of a sound-element in various words. This 
 objection, if it be any, is untenable. Paul's theory has in its 
 favor not only, as has been said, the probability of a plausible 
 a posteriori explanation, but it is supported by clinical experi- 
 ence as well as by general psychological arguments. It is a 
 fact of common medical observation that a lesion of the brain 
 may destroy or impair the ability to pronounce some certain 
 sound or sound-group. And when thus either the sound- 
 picture or the necessary sensation of motion, or only the correct 
 coordinative regulation of the different motions implied, is lost, 
 the respective sounds will drop out or suffer mutilation in all 
 words in which they would normally occur. From the point 
 of view of psychology, on the other hand, the explanation 
 here adopted is apparently included in the theory of apper- 
 ception and association, and only those can consistently object 
 to it, as to almost everything else here proferred, who are not 
 ready to accept the association theory itself. Our discus- 
 sion thus leads us in a somewhat different way to the same 
 result which Paul obtained — to the law-theory. For, as we see 
 it now, it is not a number of individual sounds in so many 
 
24 G. E, KARSTEN. 
 
 words which change, but it is one sound-picture entering into 
 the formation of so many word-pictures, that changes. 
 
 The fact that the process may be at any time interrupted 
 and cut short in individual words, when the latter undergo 
 the influence of other factors, does of course not in the least 
 alter the psychological aspect of the matter. Nor does it 
 diminish the importance of the law-theory either from a 
 philological or from a psychological stand-point. A few hints 
 may here suffice : upon the law-theory depends, in historical 
 philology our decision as to the causes of innumerable devi- 
 ations from what seems to be the normal development of 
 sounds ; while psychology may hope some day, when both the 
 genesis and the acoustics of sounds will be better under- 
 stood, to ascertain from the historical sound-change the exact 
 extent of individual sound- varieties that may be apperceived 
 into one sound-picture. 
 
 We need to add but little to explain the nature of Analogy. 
 As apperception is related to association — see page 319 — , so 
 Phonetic Law is related to Analogy. The underlying motive 
 is the same, but the materials and consequently the results are 
 different. Phonetic Law is based upon the union of practically 
 equal sounds into one and the same sound-picture ; this involves 
 sameness of fate. Analogy rests upon the association between 
 different sound-pictures or word-pictures which have something 
 in common with each other. Here, then, the equal uniting 
 parts alone are apperceived into the same memory-picture, 
 while the remainders of the larger, more complex speech-units 
 form each an impression of its own. For instance, there are 
 no two suffixes or words in any language which resemble each 
 other so much in form and in meaning as the o in hone resembles 
 the in stone and which, therefore, could be reflected by one 
 single picture in our memory. Each prefix, suffix, ending, 
 word, has in our mind a representative, a memory-picture of 
 its own, and this composite picture comprises the various 
 occurences of the prefix in different words, or of the word in 
 different sentences, as one and the same sound-picture covers 
 
PHONETIC LAW AND ANALOGY. 25 
 
 the corresponding sounds as they enter into the formation of 
 various words. All the different memory-pictures, however, 
 are, directly or indirectly, connected with each other by some 
 similarity, and in proportion as they are equal, must a change 
 of one affect the others also. Now the degree of similarity 
 varies greatly, and again, one and the same element of equality 
 may be common to a larger or smaller number of more complex 
 organisms. Thus, groups are formed on the basis of the 
 element of equality, and, as is natural in the manifold inter- 
 relations between diverse complicated units, the various associ- 
 ations may conflict with each other in their influence upon the 
 speech -material. We will first consider the simplest kind 
 of analogy, namely, that between the different innervations 
 and sound-pictures. 
 
 The innervations and sound-pictures which, so far, we have 
 treated as units, are at the same time of a complex nature. 
 Every sensation of motion necessary to produce a speech-sound, 
 governs the movement of all the organs of the vocal apparatus 
 individually and also regulates the correct coordination of these 
 movements; and on the other hand, every sound-picture 
 comprises the results of complicated vibrations. Now, various 
 innervations may and actually do resemble each other in some 
 features, and every particle which any two or more sound- 
 innervations have in common, is of course the identical inner- 
 vation of its own wherever it occurs ; a change which affects 
 it, must modify all the sound-innervations concerned in a 
 parallel manner ; and so with sound-pictures. There can be 
 no doubt that this phonetic analogy^ or better perhaps this 
 anaphonyy^ is the main source of the peculiar harmony of 
 sounds and sound-changes which — as Sievers first observed — 
 shows itself within individual languages. The other source 
 is probably to be found in the common basis of articulation, 
 the result of the sum total of the individuaPs speech peculiari- 
 
 ^ We suggest the word " anaphony " for this phenomenon, in order that 
 the expression " phonetic analogy " may be reserved for those interrelations 
 between word-dements which are based upon eijuality of sounds. 
 
26 . G. E. KAESTEN. 
 
 ties. Thus we understand the unmistakable parallelism of the 
 various sound changes comprised under the name of Grimm's 
 Law, or the development of voiceless final consonants in Gothic, 
 Old Norse, or Old High German ; the unrounding of palatalized 
 vowels in German dialects ; the far-reaching phenomenon of 
 i and it-umlaut in Germanics ; the opening and vocalizing — 
 " das Stimmhaftwerden " — of intervocal Latin consonants, or 
 the higher oif-glide of long, narrow vowels — o : ou, e: ei — and 
 the higher on-glide of the corresponding wide vowels — o : uo, 
 e : ie — in Old French ; in short, all those larger movements 
 which modern phonology more and more clearly recognizes. 
 
 In the interrelations between larger speech-elements — sylla- 
 bles, words, sentences — exactly the same motives are at work. 
 In order to recognize this fact it is above all necessary to 
 avoid a mistake which is apt to beset the mind of the philolo- 
 gist whose attention is directed principally towards changes 
 of word-forms. The student of historical phonology finds the 
 word analogy particularly applied to words, which show an 
 irregularity in their phonetic development. Formerly these 
 word-forms were attributed to false analogy, and that expres- 
 sion — while it seems to ascribe an undue superiority to forms 
 sanctioned by usage — had at least the advantage of clearly 
 indicating to the beginner, that there is also such a thing as 
 legitimate analogy at work in Language. To be sure it was 
 this very idea, that an analogy was never a false one in itself, 
 which caused Paul and others to protest against the odious 
 adjective. 
 
 However, as the word analogy seems, in the minds of many, 
 to imply the idea of something irregular, exceptional, it may 
 be well to emphasize the fact that Analogy is the main consti- 
 tutive factor in linguistic morphology and in syntax. To it 
 are due our systems of declension, conjugation, comparison, 
 etc. ; without it the individual would be strictly limited to the 
 mechanical reproduction of such word-forms, as he has actually 
 heard, while through analogy we are able to form new words 
 which we never heard, by combining known elements after 
 
PHONETIC LAW AND ANALOGY. 27 
 
 the pattern of other known combinations ; by analogy we speak 
 in sentences. When the child has heard some comparatives 
 ending in -er as against some positives without that ending, a 
 memory-picture is unconsciously formed containing the sound 
 group -er, as the ending of an adjective, connected with the 
 idea of comparative. This memory-picture presents itself, 
 whenever either of its parts is touched upon ; the ending -er is 
 added to adjectives when the idea of comparative agitates the 
 mind, and vice versa. Similarly, after having heard and under- 
 stood the structure of a more complicated statement, a sentence, 
 the mind connects the logical categories and the interrelations 
 of the same with their expressions in language forms ; memory- 
 pictures of these interrelations are developed which will 
 henceforth regulate the structure of sentences equally pro- 
 portioned. Just as in reading a word which we never heard 
 our memory-pictures of sound and of innervation enable us 
 to pronounce the individual sounds. This will illustrate the 
 fact that analogy is not essentially diiferent from Phonetic 
 Law, so far as the mental process is concerned : some certain 
 speech-element occurs repeatedly in the spoken language within 
 larger speech-units; its individual occurrences — though slightly 
 differing from each other — form in our mind one composite 
 impression, one memory-picture, which regulates them all, but 
 is in turn apt to be modified by every new utterance of the 
 respective element. The result is a peculiar harmony in the 
 phonology and morphology of individual languages. 
 
 What tends to obfuscate in our mind this identity of process, 
 is the fact that analogy, so called, works with much more 
 complex organisms, and quite especially the fact that in 
 Phonetic Law we speak of equal elements — speech-sounds — 
 alone, as whole units, while in analogy we treat the much more 
 complicated elements of equality as merely parts of other 
 units, for example, we compare words when their suffix alone 
 is the same, and from the point of view of these larger units 
 we can of course only state similarity, association, whereas in 
 Phonetic Law we speak of the equal elements and state 
 
28 G. E. KAESTEN. 
 
 apperception into the same memory-picture. This may be of 
 advantage from a descriptive point of view, but it is incon- 
 sistent. Speech sounds are no more or less units of their own, 
 to be studied by themselves, than are prefixes, endings, words, 
 logical categories and their inter-relations. When this is kept 
 in mind, the fundamental identity between the two phenomena 
 will appear clearly. 
 
 The question now arises, why should the study of results, 
 historical philology, reveal so many cases of exceptions to this 
 general harmony? Why should analogy and phonetic law 
 ever counteract each other, when they are indeed manifestations 
 of the same mental process ? The answer is this : sound- 
 pictures are on the one hand connected with ideas ; they enter 
 into larger, complex units which comprise form and meaning ; 
 for example, the suffix -er, as we saw above, combines with 
 the idea of comparative, we develop the memory-picture : (idea 
 of comparative and suffix -er). On the other hand, however, 
 sound-pictures are the results of articulations, and the latter, 
 with their sensations of motion, undergo unavoidable changes 
 in accordance with the underlying phonetic conditions. Thus 
 a sound will change into a certain other sound under some 
 conditions, while under other circumstances it is modified 
 differently. Now it may and does happen, that one and the 
 same sound in the same word or word-element — that is, con- 
 nected with the same meaning — occurs under different phonetic 
 conditions, according to accent and surroundings, and thus is 
 subject to various sound-changes. In many English dialects, 
 for instance, r remains before vowels, while before consonants 
 and in pause it disappears. This sound-change which turns bird 
 into hddd, while it leaves America intact, affects our suffix -er in 
 both ways ; before words beginning with a vowel it is -er, or 
 rather -ar, before consonants and in pause it becomes -a. Both 
 suffixces are equivalent in meaning; both connect with the 
 idea of the comparative, and instead of the former unit we have 
 now two sensations ; (idea of comparative -1- suffix -ar) and 
 (idea of comparative + suffix a). 
 
PHONETIC LAW AND ANALOGY. 29 
 
 In case the idea of the comparative now agitates the mind of 
 a speaker, which of these two memory-pictures will present 
 itself? the one with -dr or the one containing -a ? Of course, 
 in the minds of the persons in whom the change, the real 
 transition^ originally took place, the sensation of motion 
 coupled with the sound-picture regulates the pronunciation in 
 accordance with the phonetic conditions; but, the change 
 having once become a fact, the original conditions are no longer 
 in every single speaker again and again the source of the same 
 change ; those conditions may no longer exist, but the result, 
 that there are now two sounds where formerly there used to 
 be one, remains, perpetuated through the transmission of two 
 ready-made sound -pictures to every new hearer. These two 
 sounds are perceived successively, as they occur in the language : 
 the child hears hddd, and it hears America^ but it may hear 
 comparatives ending in -ar and in -9. We suppose for the 
 sake of illustration, that the d -variety should have first 
 presented itself, alone or in overwhelming majority ; then it 
 alone will form a distinct memory-picture and the sensation : 
 (idea of comparative + suffix -a) will cause the child to form 
 new comparatives in -a, regardless of the following sound. 
 Of course, a man who only knows the comparative ending -9, 
 can only say heb is heb ; he could not before is insert an r 
 any more than any other consonant. 
 
 Later, however, both suffixes, -a and -ar, are introduced 
 again and again and some sort of discrimination will develop. 
 The latter may be more or less correct, that is, in- accordance 
 with the original state of affairs ; from incomplete materials, 
 unevenly introduced, of course, no just criterion can be derived. 
 Perhaps a vast majority of forms in -ar, or the fact that these 
 forms were more intensely apperceived, may create the sensation 
 that -dr is altogether preferable and the result would now be : 
 hetdr is hebr. 
 
 ^ We would here quite especially refer to Paul who first established this 
 very important distinction between change and interchange of sounds — 
 "Lautwandel" and " Lautwechsel." 
 
30 G. E. KAKSTEN. 
 
 Gradually, however, the correct sensation as to the proportion 
 between the suffixes and their phonetic conditions develops and 
 the man now says : hebr is betd, in accordance with the original 
 sources of both varieties. The sensation of this proportion 
 may even, then, affect words which do not originally belong 
 to this category, having etymologically the suffix -a exclusively, 
 the proportion " suffix -9 : following consonant = suffix -ar .•■ 
 following vowel " may produce a result : the idea was = the 
 idear — is. But even, when the right sensation of the alterna- 
 tion between -ar and -a is already developed, either one of the 
 suffixes may at times be so predominant that the respective 
 memory-picture presents itself with more ease than the sen- 
 sation of proportion, and the result may be a form which 
 from the descriptive point of view is called an exception. 
 
 This, then, is the nature of the struggle that may arise 
 between Phonetic Law and Analogy and we see plainly how 
 there can be a variety of results coming from a perfectly con- 
 sistent working of the same psychological motives, when the 
 speech-material varies that is acquired by the speaker and can 
 alone be active in him. 
 
 We refrain from going into further detail, as Paul in his 
 Frindpien and more fully Wheeler in his excellent monograph 
 have attended to the subject of Analogy so thoroughly that 
 we have nothing new to add. 
 
 GusTAF E. Karsten. 
 
QNIVERSITY OF CALIFORNIA LIBRARY