RELATIONS OF COMPARATIVE GRAMMAR 39 



of logic. But a daughter of philosophy, modern psychology, has taken 

 its place and established itself in a relation with the historical study 

 of language which is as vital and as fruitful of the best results to 

 both sciences as the old relation was artificial and barren of anything 

 but vague speculations which only disguised the ignorance of the 

 time as to linguistic development. One of the chief characteristics 

 of the language-study of the last fifty years is the increased attention 

 paid to the psychological factors in language, and never has the 

 relation between linguistics and psychology been so close as at the 

 present moment. There is no better external evidence of this than 

 the two large volumes which one of the most eminent psychologists 

 has devoted to the psychology of language and the attention which 

 has been given them by students of language, or the numerous 

 special investigations of problems in language psychology, whether 

 written by one who is primarily a comparative grammarian or by one 

 who is primarily a psychologist, or, as in some cases, under the joint 

 authorship of a representative of each science. 



In one sense all linguistic phenomena are psychological. Even 

 the regular phonetic changes which we have treated as involving 

 physiological relations have of course their psychological back- 

 ground, are, in other words, psycho- physical. 1 



But the historian of language is constantly dealing with matters 

 which involve purely psychological factors. Language is a register 

 of associations on the grandest scale. One of the most important 

 functions of the general comparative grammarian is to compare 

 the distinctions and relations which find expression in the gram- 

 matical categories of different groups of languages. These gram- 

 matical categories show the various ways in which objects and their 

 relations group themselves in the minds of different peoples. What in 

 one language is an important grammatical distinction may be ignored 

 in another. For example, gender, which plays such a r61e in our own 

 family of languages, follows only one of the many lines of division 

 between objects which find grammatical expression in this or that 

 language, such as between objects animate and inanimate, human 

 and non-human, high or low in rank, beneficent or otherwise. Again, 



1 In certain classes of phonetic changes the psychological element seems to be 

 the more obvious factor, notably in the assimilation, dissimilation, or metathesis 

 of non-contiguous sounds, which are most common in rapid or careless speech 

 and in a state of fatigue, and which are essentially pathological, momentary 

 lapses due to imperfect attention, only occasionally gaining general currency. Or 

 since such changes are by far the most common in the cases of liquids I and r 

 (e.g. marble from French marbre, Latin marmor, pilgrim from late Latin pelegrinus, 

 earlier peregrinus, etc.; in New Orleans one hears a certain confection called 

 indifferently praline or plarine), shall we not rather say that the physical relation- 

 ship of these sounds in their formation is such as to require greater attention 

 than other sounds for their proper adjustment to one another, so that even here 

 the physical element is equally fundamental? The question at least illustrates the 

 impossibility of separating the factors sharply. 



