Wikipedia:Reference desk/Archives/Language/2021 May 10
Language desk | ||
---|---|---|
< May 9 | << Apr | May | Jun >> | Current desk > |
Welcome to the Wikipedia Language Reference Desk Archives |
---|
The page you are currently viewing is a transcluded archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages. |
May 10
editUncertainty about row and column headers
editThis question is mainly about interpreting an English sentence, but insight on whistled languages is also appreciated.
Whistled_language#Techniques had this table:
i | e | a | o | u | |
---|---|---|---|---|---|
i | 15 | 1 | |||
e | 1 | 1 | |||
a | 79 | 5 | |||
o | 4 | 15 | 3 | ||
u | 2 | 2 |
with a note
- Confusion matrix of the vowels in the perception test. 'Produced' vowels are displayed horizontally and 'perceived' vowels vertically (Numbers in bold correspond to correct identifications).
Produced Perceived |
i | e | a | o | u |
---|---|---|---|---|---|
i | 15 | 1 | |||
e | 1 | 1 | |||
a | 79 | 5 | |||
o | 4 | 15 | 3 | ||
u | 2 | 2 |
I've labelled the column and row headers to make it clearer to the reader. Did I get it right, or is it the other way around?
Thanks,
cmɢʟee⎆τaʟκ 22:14, 10 May 2021 (UTC)
- Oh wow, I didn't know we could do that. —Tamfang (talk) 00:42, 13 May 2021 (UTC)
- This matrix is almost the transpose of the matrix from the source, except that the latter has the value 3 at position [a,i] (corresponding to the position of the value 1 in the top row of the above matrices). The source does not specify the meaning of the rows versus the columns, but provides sufficient indications. In the same appendix giving these matrices, there is a table of 68 whistled/perceived pairs of VCV triples. There are four cases of a misidentification involving an /i/: (30) iti – ati; (37) ipi – ati; (43) ete – ati; (55) iti – ati. So there are three cases of a produced /i/ being perceived as an /a/, but none the other way around. Conversely, there is one case of a produced /e/ being perceived as an /i/, but none the other way around. I cannot exactly reproduce the matrix in the source from the table of pairs of triples, but what I get is very close:
h\w| i e a o u ---+-------------------- i | 15 1 e | 1 a | 3 1 76 4 o | 5 15 2 u | 1 2 ? | 2 6 2
- Here, "h" stands for "heard" (= perceived), and "w" stands for "whistled" (= produced). So it looks like the corner label pair Perceived\Produced is the wrong way around. --Lambiam 10:09, 11 May 2021 (UTC)
- @Lambiam: Thank you so much for investigating it and presenting your findings clearly. I'll update the table accordingly. Cheers, cmɢʟee⎆τaʟκ 18:26, 13 May 2021 (UTC)