For those who think that irony "is almost always indicated by tone of voice", a little quiz:
Which of these are literal (positive evaluation of something) and which are ironic (negative evaluation of something)?
Answers tomorrow The answers are below. (I've re-opened comments but no more responses will be tallied, so please don't bother giving them to us — the comments are open for discussion of what the exercise means, if anything…)
First, let me point out that this was a poorly-run experiment. I should have set it up so that listeners are exposed to one clip at a time, and don't get to see other people's responses, and have to make a forced choice. I understand that Christoph Draxler has written some neat HTML5 code for running perception experiments, and plans to make it available to others, so at some point in the not-too-distant future we'll be able to do proper perception experiments (leaving aside issues of variation in listening technology, acoustic environment, and listener seriousness…).
For the purposes of this unscientific straw poll, I added up the responses anyhow, assigning half a vote in each direction when people said they couldn't tell about a given stimulus. Some people didn't respond at all with respect to some items, and I left those out.
The table below gives the tally, the percentages, and the truth (which is in each case obvious from the (textual) context).
|1||22.5 (59%)||15.5 (41%)||ironic|
|2||6.5 (17%)||31.5 (83%)||literal|
|3||31 (86%)||5 (14%)||literal|
|4||9.5 (25%)||28.5 (75%)||ironic|
|5||23 (61%)||15 (39%)||literal|
|6||29.5 (78%)||8.5 (22%)||literal|
If we tally up the responses for the four literal instances of "oh great", we get 90 "literal" responses and 60 "ironic" responses, so listeners were 90/150 = 60% correct.
Tallying up the responses for the two ironic instances, we get 32 "literal" responses and 44 "ironic" responses, so listeners were 44/76 = 58% correct.
Overall, listeners were 59% correct.
So the performances are apparently conveying some information about the literal/ironic distinction: but it's far from perfect. And some individual instances (especially #2) are consistently misunderstood — which means that these percentages are likely to be rather unstable, and would probably change in a better experiment with a larger number of examples.
I interpret these results as follows. The performances convey a modest amount of attitudinal (and other) information, which some listeners are using to reason (consciously or unconsciously) about likely utterance contexts and speaker attitudes, and to make a guess about whether the utterance was literal or ironic. Another way to look at the process is that listeners are responding to an ensemble of performance features that they associate to one extent or another with "irony" or "sarcasm" — perhaps even in the specific case of this common phrase. Under either way of looking at it, the performance features in question do not include anything worth calling a "sarcastic/ironic intonation" or "sarcastic/ironic tone of voice", since the same features would play a role in the evaluation of arousal, affect valence, and so on.
Let me note in passing that there's a striking contrast here between the "text" and its performance in the quantity and quality of information conveyed. Despite the indifferent acoustic quality (these clips are taken from telephone conversations), there's no question in these examples about what the words are; but whatever the intonation, tone of voice, etc., are conveying, it's much more diffuse and there's much less agreement about what it is.
These examples were picked in the following way: I searched the LDC's published telephone-conversation corpora for the sequence "oh great"; clicked on a small more-or-less random sample; and took the first six where the "oh great" part was not too much overlapped with other speech or otherwise acoustically unsuitable. This required listening to 8 examples.
Here are the textual contexts for the clips that I chose:
A: and uh as we were doing it too everybody noticed that there were all these big guys of assorted nationalities to play basketball waiting ar- for us and as we walked out i was limping and i'm like oh great i have purple tights on you know
B: uh one of the um times that we had sort of a family reunion was when uh was this i think it was last last winter we uh we went up to New Hampshire to go skiing
A: oh great
B: and uh a bunch of us got together
B: uh well i i like i said i've done mechanics all my life and i do that with the company so uh i'm a senior engineer for them so i am mechanics and i i just uh i enjoy it i really it's like a diversion for me i i think if i had to do it
A: oh great
B: if i had to do automotive repair or or anything i love to do woodwork and too and it's a diversion because i know i'd really don't have to do this okay so i think i like that because i know i don't have to do this but if if i were forced to do it that may be a different situation
A: so definitely no pets
in fact i just talked to my um husband we're supposed to visit my sister in law next week and she has a cat
and i'm thinking oh great i'm you know i need to be able to breath
B: but you know at at Texoma it's such a big lake and we don't have a boat but we're on the dock and people come in there's a lot of sandy islands out in the middle of Texoma
B: and they go out and fish around there and people come in with these ice chests just full of of you know twelve fourteen sixteen eighteen inch long fish
A: oh great
B: it's the darndest thing i've ever seen and i don't know we we're probably going to have to get us a boat one of these days soon and go out and see go out and see what's really out there
B: yeah yeah well what's your major in school
A: it's art
A: yeah fine art
B: what type of art
A: fine art
B: oh great
B: (( well that's good well can't you uh make your own greeting cards ))