Language and Folklore (Supplement). 553 
а phonetic rendering of what Г, with ears alert, could sense from the 
lips and voice of the native. How, then, explain these variations? 
My principle in deciphering the often all too slurred words 
(sounds) of the phonograph, was the same as in direct dictation; to 
note down phonetically my auricular impression without permitting 
myself to be disturbed by the semantic side of the matter, or by the 
forms of Eskimo grammar in any other way than to observe the 
necessary separation between each word or each word body. Here 
indeed the peculiar characteristic of the language tricked me so that 
I was not always able to distinguish the separate forms or to collect 
those which belonged together. For instance, I obtained two variants 
(Phon. and DD) of no. 188, for Kooitse sang the song into the phono- 
graph and dictated the text to me. While I was about to decipher 
the phonographic record I had no recollection of the DD text and 
followed only my ear in the lines here given to the left of the line: 
Phon. ilayeni åt apa nai = ilane ninätsärpa na DD 
» akon'o‘ni „dtapa'nai = akupuniya ninätsärp'a ma » 
However it was the DD text that showed the right morphological 
separation of the words and thus enabled me to understand the lines. 
If I had reread the DD text first, my deciphering of the phonograph 
would doubtless have given another result, — in reality less valuable, 
because influenced by prejudice: unprejudiced recording alone, has 
scientific value. 
As a rule, then, what sounded ambiguous in my ear Was repeated 
ambiguously in my notes (f. inst. м ) and what was clearly under- 
stood was recorded exactly as I (my ear) had sensed it, even though, in 
that particular form, it was in opposition to the usage of the language. 
The following are the principal sources of error in using a 
phonograph: 
1) The individual peculiarities (“errors”) of the informant, or 
singer, either in his pronunciation, or in the forms of speech deviating 
from the normal. 
2) The informant’s facultative voice-quality (strength and tempo) 
depending upon his state of mind or physical condition (nervousness, 
shyness, exaggerated pronunciation, exaggerated expression, remodel- 
ling or changing, omissions, weak presentation, etc.) 
3) The informants grade of consciousness that his record should 
be arranged, his pronunciation, or the content of his narrative sim- 
plified for the sake of a foreigner’s better understanding, 
4) A phonograph’s characteristics in general (lack of clearness in 
differentiating between the fenues, or between fenues and medic, or 
between open or breathed consonants etc.) 
