Conversational Grunts cause serious problems for automatic voice-transcribing algorithms. In fact, as late as year 2000, experts were writing : “… no satisfactory scheme for transcribing these items exists.” The quote is from Nigel Ward, who is currently professor of computer science at The university of Texas at El Paso. The professor has been analysing, cataloguing and generally investigating Conversational Grunts for more than 15 years.
Some examples of his grunt-based publications include :
The Relationship between Sound and Meaning in Japanese Back-channel Grunts. (4th Meeting of the (Japanese) Association for Natural Language Processing, 1998.)
Issues in the Transcription of English Conversational Grunts (1st SIGdial Workshop on Discourse and Dialogue. ACL. 2000)
Non-Lexical Conversational Sounds in American English Pragmatics & Cognition 14:1. 2006. 212 pp. (pp. 129–182)
A Model of Conversational Grunts in American English (2002) (IN PAPERS FROM THE 6TH REGIONAL MEETING)
BONUS [1]
A substantial directory of Conversational Grunts, including sound files [in .au format] can be found here via the professor’s website.
BONUS [2]
A glossary of 100 common grunts (from ‘Issues in the Transcription of English Conversational Grunts’)
[clear-throat]
tsk
tsk-naa
tsk-neeu
tsk-ooh
tsk-yeah
[inhale]
[unsticking]
aa
achh
ah
ahh
ai
am
ao
aoo
aum
eah
ehh
h-nmm
haah
hh
hh-aaaah
hhh
hhh-uuuh
hhn
hmm
hmmmmm
hn
hn-hn
huh
i
iiyeah
m-hm
mm
mm-hm
mm-mm
mmm
myeah
nn-hn
nn-nnn
nu
nuuuuu
nyaa-haao
nyeah
o-w
oa
oh
oh-eh
oh-kay
oh-okay
oh-yeah
okay
okay-hh
ooa
ookay
oooh
ooooh
oop-ep-oop
u-kay
u-uh
u-uun
uam
uh
uh-hn
uh-hn-uh-hn
uh-huh
uh-mm
uh-uh
uh-uhmmm
uhh
uhhm
ukay
um
um-hm-uh-hm
umm
ummum
un-hn
unkay
unununu
uu
uum
unmm
uun
uuuh
uuuuuuu
wow
yah-yeah
ye
yeah
yeah-okay
yeah-yeah
yeahaah
yeahh
yegh
yeh-yeah
yei
yo
yyeah
[apologies for any transcription errors, ahem, M.G.]