Page:Mind (Old Series) Volume 12.djvu/88

 EXPERIMENTS ON "PREHENSION". By Joseph Jacobs. It is obvious that there is a limit to the power of reproducing sounds accurately. Anyone can say Bo after once hearing it : few could catch the name of the Greek statesman M. Papa- michalopoulos without the need of a repetition. It is here attempted to ascertain the normal limits of such reproduc- tion in various circumstances and under varying conditions. At first experiments were made with nonsense-syllables like cral- forg-mul-tal-nop, as suggested by Ebbinghaus's experiments. It was found, however, that the syllables used varied greatly in relative difficulty of pronunciation and in relative facility of rhythm. After a few trials they were abandoned for letters (omitting "double u") and numerals (omitting "seven" as dis- syllabic). It was found on the whole that the facility of repro- ducing the different kinds of sounds, after once hearing them, went together in a tolerably constant ratio. Thus a number of school- girls who could repeat on an average 6*1 nonsense-syllables could repeat 7-3 letters and 9'3 numerals. The explanation for this order of difficulty is not far to seek. The syllables, as contrasted with numerals and letters, are new to the hearer, have to be learnt, and absorb more energy ; then, again, their grotesqueness would distract the attention more. The comparative diffi- culty of reproducing letters as compared with numerals is not so obvious. Reading accustoms us to take letters in groups having a phonetic value, and collocations of consonants like bsvlrtm strike us in a minor degree with the same sense of incongruity which prevents our minds from easily assimilating a conjunction like dak-mil-tak-Un-roz. Numerals, on the other hand, have few, if any, associations of contiguity, and we are accustomed to find them in haphazard order. Finally, our expectant attention has only to search among nine numerals, whereas it has to be ready to select from twenty-five letters. School-habits, however, might modify these conditions, and the cases were not infrequent in which the limit for letters was higher than for numerals : thus in one set of schoolboys no less than 14 boys out of 88 could repeat more letters than they could numerals, while 33 of the remainder had the same limit for both.