Wednesday, August 20, 2008

regrets, I've had a few

Anatol Stefanowitsch at Bremer Sprachblog takes another look at a classic xkcd


(pop-up: And nothing for I'm glad I saw 'Epic Movie')

AS:

Ich bin ja ein erklärter XKCD-Fan, aber ich finde, dass er in diesem Fall durch die Zusammenfassung der Ergebnisse für him und her eine Chance vertan hat, detaillierteres Wissen über die menschliche Natur herauszufinden. (I'm an avowed xkcd fan, but I think in this case, by amalgamating the results for him and her, he lost a chance to bring to light a more detailed knowledge of human nature.)

AS makes good this important omission, first by running searches in English according to gender:

(pop-up: Tu es oder tu es nicht, du wirst beide bereuen (Sören Kierkegaard) [do it or don't do it, you'll regret both])

and THEN by running searches for Ich hätte ihn/sie (nicht) küssen sollen


(pop-up: Das einzige, was ich bereue, ist das ich nicht jemand anders bin (Woody Allen) [the only think I regret is that I'm not somebody else])

Stefanowitsch has revealed several striking disparities between the anglophone and germanophone worlds of sexual regret - or, at least, disparities between the pools of regretful English and German speakers who feel called upon to share their disappointment on the WWW. (Perhaps English-speaking men are likelier than both English-speaking women and German-speaking men AND women to have no alternative to the kindness of strangers.)

There's a lesson to be learnt.

The lesson, of course, is that Randall Monroe and Antaol Stefanowitsch should collaborate on a T-shirt. Guys. Guys. You know it makes sense.

(Bremer Sprachblog, as so often, brings to mind the possibly apocryphal British headline: Fog in Channel: Continent cut off. In this case, it's the non-German-speaking world that's cut off from this consistently excellent blog; if you know any German at all, check out the rest here.)

9 comments:

ghost said...

NOt sure if it's not possible that Germans would be less likely, in online discourse, to use that very sentence, especially in light of the fact that Germans writing online use heaps of english (or english'd) words, and abbreviations. Also, possibly Germans are less likely to use "ihn/sie" instead of "den Typ" or "Peter". NOt to mention that none of the results points to "Germans" but only to "people using the German language at least once when writing online", and there's a major difference.

Helen DeWitt said...

There are people whose native tongue is German in Germany, Austria, Switzerland, Lichtenstein, Poland, and (I gather) Texas, as well as (no doubt) any number of places that don't instantly come to mind. While some of them may use an English sentence to express regret, it's hard to believe that they are not less likely to do so than native speakers of English.

The point to which I was gesturing was that we should be wary of the Dewey effect. In the run-up to the Dewey-Truman presidential election, a telephone poll was conducted in which Dewey appeared to be sure of a landslide. The pollsters had failed to notice that the poll's results were weighted in favour of voters who had a telephone. It's entirely possible that English-speaking women use the sentence "I should have kissed him" as often as English-speaking men use the sentence "I should have kissd her" - but that they do so much more than men in text messages, conversations on cellphones, and conversations in the physical presence of their friends. Similar considerations may come into play if we compare the corpus of German sentences with that of English sentences. (These factors may equally well, of course, be insignificant, but there's no way of finding that out from a Google search.)

I take it that anyone who uses the German language online can be assumed to be a speaker of German. If anyone has managed to generate German sentences without speaking the language, however, it seems unlikely that these sentences would represent a significant proportion of the relevant corpus of German sentences.

Anonymous said...

One other quibble. I take the implied message of the XKCD graph to be that, at a rate of about 10:1, you are more likely to regret not doing something than you are to do it, so be bold and start giving out the smooches. But all we can really say about it is that the number of people who regret not kissing him/her is greater than the number of those who regret doing it. To begin drawing conclusions about these numbers, we'd need to know the total number of people who did and did not kiss him/her. Then we could establish some sort of probability of regret based on the rate of kissing/not kissing within the general population.

And then we can start handing out the worry-free smooches.

Levi Stahl said...

This is the second illicit-kissing-related item I've seen this week, the first being a line that Ed Park quoted from Joe Meno's "An Apple Could Make You Laugh":

"We do not actually do any work in the office anymore, other than trying to imagine what it would be like to kiss each other. We have been thinking about it so long we have forgotten what it is we should be doing."

Clearly something, possibly regret-causing, is in the air.

Cecilio Morales said...

As to the various disclaimers regarding German-speakers ... feh! The same could be said of English-speakers insofar as usage and national variety.

Other biases: the author of xkcd is a guy; searching for something doesn't mean agreeing with it.

That said, I love xkcd and smooching. The shoulds have it, the hell with what Germans do.

"Post-Google" by TAR ART RAT said...

on that note, I can't remember if I have sent this along already or not, but the Graphing Songs never cease to amuse me

SeanONeill said...

Um, no.
As another commenter elsewhere has pointed out:
i should have kissed him OR her = 11,200,000

i should not have kissed him OR her = 4,140,000
i shouldnt have kissed him OR her = 641,000
i should'nt have kissed him OR her = 10,400
i shouldn't have kissed him OR her = 3,330,000

thus should not = 4m + 0.6m + 0.01m + 3.3m = 8,121,400

RESULT – more people regret kissing they did than not kissing someone they didn't but the difference is not as huge as the graph suggests...

11m v 8m ~ 1.4

Quodlibet said...

I don't think it's legitimate to infer the sex of the speaker from the sex of the person kissed or unkissed. Men do kiss men, and women women.

Helen DeWitt said...

john, you're absolutely right, this sort of carelessness is the sort of thing people haul into the Hall of Shame when pointing to the inferiority of blogs to books that get edited before publication. Sadly, publishers are happy to spend a lot of money on a copy-editor, ensuring that the text conforms to the Chicago Manual of Style, and have shown little enthusiasm for hiring a house philosopher to spot logical howlers, so the chances of getting logical howlers spotted is actually higher if one sends a post all unedited out into the world for public scrutiny. Excelsior.