THE WORDS AND SOUNDS OF TELEPHONE CONVERSATION 291 



tional phrases and it makes use of synonyms and a vocabulary more or 

 less ample according to the writer's fancy and ability. In conversation 

 attention seems to be paid more to the thought than the form of ex- 

 pression, with the exception, perhaps, that certain modes acceptable in 

 writing may be considered as too formal for conversation. It is doubt- 

 ful, however, that conversation should be described as more concise 

 than written matter. The sentences are, indeed, likely to be shorter. 

 They are often incomplete, in fact. But often in conversation even a 

 single statement is completed only after a number of fumbling attempts, 

 an oral manifestation of crystallizing thought, whereas in written 

 matter the final expression alone would appear. In repetition of a 

 thought, synonyms are less likely to be found in conversation than in 

 written matter. Dependent clauses are less frequent than in written 

 matter. Qualification and description often take the form of separate 

 sentences, so that those words characteristic of involved construction 

 tend to be less prominent in conversation, while the framework words, 

 such as the auxiliary verbs and pronouns, are more intensively used. 

 These differences, which tend to restrict the vocabulary, will be found 

 reflected in the comparisons given later in this paper. 



The material for the present study was obtained from telephone 

 conversations over typical toll circuits terminating in the city of New 

 York. The process of noting the words of the conversations was 

 carried out in the following manner: During one week the observer 

 recorded nothing but the nouns used, during another week she re- 

 corded only verbs, and during a third week only adjectives and adverbs. 

 This routine was repeated until observations had been made on 500 

 conversations for nouns, 500 conversations for verbs, and 500 con- 

 versations for adjectives and adverbs. Three other classes of words 

 were recorded: prepositions and conjunctions, pronouns, and articles; 

 but for these classes approximately 150 conversations in each case 

 were judged to be sufficient. 



Certain classes of words were, for various reasons, omitted entirely. 

 These are names, titles, exclamations, letters, numbers and the name- 

 less sound which may be transliterated as "er" or "uh," so frequently 

 punctuating a haltingly expressed sentence. A more comprehensive 

 method, but based on a much smaller number of conversations, indi- 

 cates that the ratio of the total number of occurrences of words 

 in the omitted classes to the number of occurrences of the words 

 discussed in this paper is about one to four. Within the omitted 

 group the division is roughly as follows: proper names and titles, 20 

 per cent; exclamations and interjections, such as "yes," "no," "well," 

 "yeah," "uh-huh," "oh," "all right," "hello," "good-by," laughter 



