Fatter for longer (sigh)
« previous post | next post »
Here's a doubly embarrassing confession. First it involves my use of a construction that I love to make fun of. Secondly my spontaneously generated example is unfortunately also a true sentence.
I was trying on four dresses that have been stored in the attic for a while to see if I could avoid having to shop for a formal dress in Chicago on Friday for the Friday black tie dinner that precedes the Saturday honorary doctorate. I didn't think I was going to be able to fit into any of them, since I've gained back all the weight I lost around 2008-9 and am now close to an all-time maximum. But to my in some ways happy surprise, I found that I could sort of fit into two of them, including the best one. And my surprise was expressed (just talking silently to myself, but obviously in real sentences, since this sentence immediately caught my attention as soon as I "said" it) as "Gosh, I've been fatter for longer than I thought". (The happy part is I may not have to go shopping on Friday, or at least it won't be obligatory to buy a new dress, which takes off the pressure that accompanies last-minute obligatory shopping.)
I still reject that sentence, even though I said it .
I consider most sentences with two 'more's and one 'than' to be either semantically anomalous or else ambiguous among a bunch of readings that could be generated involving various sorts of ellipsis. I think clever advertisers love them and often use them, because they SOUND stronger than a sentence with just one 'more', but to try to pin down a claim of false advertising would be really hard because what they mean is much less clear. (From around 1958: More people get more satisfaction out of L&M than (out of) any other cigarette. From H&R Block signs on their stores in recent decades: "We get bigger refunds for more people."
I know (I think) what I was thinking when I 'said' that sentence, even though I can't see any way to get this as the actual meaning by any compositional semantic rules I know of: [Over all this time that I've been overweight,] I guess I've been up toward my highest weight for more of that time than I realized.
It's not the same as "I've been fat for longer than I thought" — I know that I've been overweight pretty much ever since I quit smoking in about 1986, the only real exception being for a period around 2008-10 when I really worked on it. The "fatter" rather than "fat" has to do with being up in the upper range of my overweight range.
And this definitely isn't an example that could be analyzed via conjunction as some double-'more' sentences seem to be. [One reading of "John put more marbles in more boxes than anyone else" is said to be "J put more marbles in boxes than anyone else and he put marbles in more boxes than anyone else"] Because it's false that "I've been fatter than I thought" — I know how fat I've been — and it's false that "I've been fat for longer than I thought" for reasons given above. So it's certainly not the conjunction of those two, since I wasn't asserting either one.
Oh, I hate this post! But it's all true and it's probably good for me to say it. I hope it will motivate me to get myself back into better shape again. And I think the endless interestingness of this problematic but frequent construction is worth a little embarrassment.
Ø said,
June 11, 2014 @ 7:08 pm
I don't hear your sentence as being seriously ambiguous. I see _much_ more ambiguity in things like "last best chance".
Avery Andrews said,
June 11, 2014 @ 7:48 pm
I find Barbara's example rather difficult, for various reasons, but a simple one I suggest is:
John did a crazier thing at a higher speed on the McGrath Highway than Mary did, which is I think true iff John did something scary driving on the McGrath Highway and so did Mary, and the craziness of John's think exceeded that of Mary's and also the speed (so we have componentwise comparison of ordered pairs .
(based on a sentence actually uttered in the early 1970s by Dorothy Siegel on the McGrath Highway "People do crazier things at higher speeds on the McGrath Highway than they do other places")
Joe Fineman said,
June 11, 2014 @ 8:20 pm
Does it mean "Gosh, I've been fatter than I thought, and for a longer time than I thought"? That might reasonably collapse to "Gosh, I've been fatter, and for longer, than I thought", but I wouldn't want to do without the "and".
Tom V said,
June 11, 2014 @ 9:08 pm
I read it as "I've been fat enough not to fit in all four of these dresses [or some other benchmark piece of clothing] for longer than I though I had."
Or perhaps I've lived long enough with an idealized but unreached weight that the sentence came across perfectly clear and unambiguous to me.
Barbara Partee said,
June 11, 2014 @ 10:26 pm
To Tom V —
Oh, no, it was almost the opposite. The newest of those dresses is 3 years old, the oldest more than 20. I thought I wouldn't fit into any of them, meaning I thought I had only recently gotten as much overweight as I am now. I wasn't totally shocked to find that I could sort of fit into the 3 year old one, but very surprised that I could fit into one that was quite a lot older (I don't know exactly how old). That surprise means I thought I was quite a bit thinner at the times of those dresses.
I guess a clear and well-formed sentence would be "I didn't think I was that fat that long ago." The "that long ago" is a little vague – it covers a range of different times in the past.
Barbara Partee said,
June 11, 2014 @ 10:27 pm
@Joe Fineman,
That's what I would call the conjunction type of reading, but I don't think it fits this case. I tried to say why in the body of the post.
Barbara Partee said,
June 11, 2014 @ 10:29 pm
Also, I agree that people don't often perceive any ambiguity, or even any problem, in sentences with two 'more's and one 'than', but it's a puzzle why not. They seem to allow easy 'surface processing', but when you try to work out exactly what would falsify such a sentence, it's not so easy, I think.
J.W. Brewer said,
June 12, 2014 @ 12:58 am
I had no idea that receiving an honorary degree required wardrobe choices other than donning a traditional academic gown which is typically loose enough to cover a multitude of body-image sins. But surely the comment thread is burying the lede, which is mazel tov on the recognition (and one of the other recipients is also a linguistics guy or at least a Semiticist — what the hell sort of major research university thinks language stuff is sufficiently non-marginal to devote that high a percentage of its honorary degree bandwidth to that sort of scholar?).
If it's any consolation, I expect that your male co-honorees may have experienced the same phenomenon I have, viz. that the waistband of the trousers to that tuxedo they got a decade or more back has mysteriously shrunk over time, perhaps due to unfavorable climactic conditions in the closet.
unekdoud said,
June 12, 2014 @ 3:40 am
In my interpretation, "fatter" is just "fat relative to X", where X is a baseline to be specified by the word "than". In that case, "longer" definitely means "longer than I thought", while "fatter" may just be "fatter than normal". This reasoning supports either "fatter for longer" or "fat for longer", while taking into account that "fat" in the second case may be subjective.
A construction which appears to defy this logic is Avery's example where the whole meaning of "People do crazier things at higher speeds on the McGrath Highway than they do other places" can be changed by omitting everything after "than". It's a unique sort of garden path which relies on the nouns being plural, and the truncated version uses comparatives completely differently.
Greg Hooper said,
June 12, 2014 @ 5:27 am
I had unekdoud's interpretation as well – "I have been fatter than how I imagine myself for longer than I thought" In other words 'I' need to update my body image.
Adam Funk said,
June 12, 2014 @ 5:39 am
ISAGN for a "linguists' confessions" category.
Maryellen MacDonald said,
June 12, 2014 @ 8:02 am
I think this sentence owes to a combination of two forces in language production (relevant even if the sentence isn't uttered aloud). First is in the process of mapping the nonlinguistic/prelinguistic thoughts (surprise that the dresses fit) to linguistic content. At that point, notions of more(weight) and more(time) are both relevant and can have an influence on development of the utterance. This point is related but not identical to several comments above. I'm not saying that BP said X but meant Y; I'm saying that because both "fat for longer" and "fatter back then" were both compatible with the surprise, both were briefly influencing utterance planning before she settled on the intended utterance and became consciously aware of this intended utterance.
Second, the error is an example of a bound morpheme speech error, in which -er is duplicated. These are well-attested in error corpora, and they're much more common when there's something else going on, in this case that there were two alternative notions temporarily influencing utterance planning. That is, analyses of error patterns have been used to argue for the non-independence of early stages of planning and the final assembly of the morpho-phonological form, and I see this error as an example of this sort.
A couple of observations about this argument:
1. I realize that it's enormously unsatisfying to rely on a claim that even thought the speaker has conscious awareness of intending to say X, it was nonetheless the case that both X and Y were temporarily (unconsciously) considered during early planning. I find this account plausible (though unproveable for any one example) because many psycholinguistic studies of language production are designed to test, and find support for, exactly this claim of brief influence of both X and Y despite conscious awareness of only X.
2. For what it's worth, I think that a very similar argument is going to prove useful for accounting for a phenomenon in many other posts on this blog, misnegation. That is, I think that at least some of these misnegation errors owe to the temporary viability of two ways of conveying the message, one being not(X) and the other being not(Y), with the result that sometimes the utterance comes out conveying both not(X) and not(Y). I've been thinking about ways to induce these errors and test this idea more formally. It's interesting to consider that what we might call misquantification errors like this one might receive similar treatment. Other examples?
3. I also realize that pointing to an interaction of two factors (roughly at the early planning and later phonological assembly stages of production) will be unsatisfying to someone who hopes for a single-factor account. Again I'm less bothered here because there's such good evidence in production for interactions of exactly this sort.
marie-lucie said,
June 12, 2014 @ 9:03 am
JWB, the academic gown will cover the clothing worn at the ceremony, but it will not be worn at the black tie dinner, for which a long dress is appropriate (as is a tuxedo for the male counterpart).
Don said,
June 12, 2014 @ 9:33 am
Song I listened to this morning: "I've Had Harder Times from Better Women Than You" by the Zen Blues Quartet.
JS said,
June 12, 2014 @ 9:45 am
example times/states:
(1) in 1995 I believed I was a little bit overweight
(2) in 2005 I believed I was quite overweight
(3) in 2014 i realize that i was in fact quite overweight not only in 2005 but also as early as 1995
in encoding these facts one problem seems to be the reference of "thought": is it to what I thought at time 1, or to what I thought that has only changed as of time 3, my 2014 realization?
we might simply utter:
[A] "I have been fat longer than I thought," i.e., thought until just now
this "thought" clearly refers to time 3, that is, to what I thought until the 2014 realization, but doesn't do full justice to the facts, specifically, my time 1 impression of slight fatness vs. my time 3 impression that an impression of quite fatness ought in retrospect to have commenced by time 1
whereas
[B] "I have been fatter for longer than I thought"
attempts to introduce this additional information but ends up ambiguous in its effort to mean the rather complicated
[C] "I was fatter than I thought at time 1, and time 1 is longer ago than I thought until time 3 just now."
or so my story goes
J. W. Brewer said,
June 12, 2014 @ 11:37 am
marie-lucie: Having alas thus far never been awarded an honorary doctorate myself, I simply had not previously considered that a black-tie dinner might be part of the process. Once the fact is presented it is not particularly surprising, although I guess a "business attire" dinner or even a "professorial-tweed-jacket-with-those-suede-elbow-patches" (yeah, I don't know what the female equivalent of that precise degree of formality-v.-informality would be either) dinner would also not be surprising and I would imagine practice varies from university to university.
marie-lucie said,
June 12, 2014 @ 2:10 pm
JWB, read Barbara's description of the planned festivities in the first few sentences of her post. "Black tie" is primarily directed at men, but the phrase is short for several other details of both male and female attire which are expected on relevant occasions.
Belial said,
June 12, 2014 @ 4:05 pm
Citius, altius, fortius… quam aliquis!
J. W. Brewer said,
June 12, 2014 @ 4:39 pm
m-l: I'm sorry. My previous posts had been insufficiently explicit. I had not considered prior to reading Prof. Partee's original post the need to worry about black-tie attire in connection with receiving an honorary degree – the original post made it clear enough. "Black tie" of course may impose maximally asymmetric burdens by sex, since an adult male is comparatively unlikely to have more than one suitable outfit in his closet, and when you have only one suitable outfit, there are no wardrobe choices to be made, other than maybe choosing among cuff-link options.
Mr Punch said,
June 13, 2014 @ 10:49 am
The H&R Block ad, "We get bigger refunds for more people," actually makes sense. Tax preparation services compete with each other, but first they have to persuade customers to use a service rather than preparing their own returns. The common selling point in the first phase is that the pros will get a bigger refund. In other words, "bigger refunds" are the product. If there's ambiguity, it's in the "more people" – does this mean a higher proportion of customers, or simply that Block has the most customers?
Barbara Partee said,
June 13, 2014 @ 1:29 pm
I've been too busy to follow (we've now got the tux, I've got new shoes to go with the 2011 dress that surprisingly still fits (more or less, but I didn't see anything half as nice at Macy's this morning), but I see a lot of interesting ideas in these comments. Maryellen MacDonalds remarks on processing research look very relevant — I'm not sure I even know "exactly" what I thought, so it's plausible that there were competing propositions at an unconscious level during early processing. JS's story could be relevant – I don't think I had exactly those thoughts, but maybe something similar.
To Mr Punch, I think I would argue a bit about the H&R Block case. For instance, I think one's first impression is that they are claiming that they get bigger refunds for their clients than other tax preparers do, AND that they get refunds for more people than any other tax preparer does. But if either of those claims was challenged in court, they could just bring in a linguist (like me) to argue that there is no straightforward compositional syntactic and semantic analysis that entails either of those things.
I think what you're saying is reasonable with respect to how normal people will interpret it. And what makes these things so interesting to me is the gulf between that normal impression and what anybody's syntactic and semantic analyses can account for. And I even think our analyses are right, so that it's a really interesting gulf between some sort of 'surface processing' and some deeper analysis that we could do but rarely do.
Jeff W said,
June 13, 2014 @ 5:59 pm
Well, speaking just for myself, that wasn’t my first impression. I read it as Mr Punch explained it, with “more people” meaning “proportionately more” but realizing that it could just mean a higher absolute number. I’d assume “more people” means “a greater proportion of our customers” in comparison to what the competitors get for their customers. I wouldn’t think that there was a claim that they were getting more refunds of any type for more people—in fact, I can’t even figure out a way to make the phrase read that way.
The black tie event is probably going on in just a few hours. Have a fun time and just enjoy yourself!
Bobbie said,
June 14, 2014 @ 6:49 pm
I have gained less than one pound per year since I graduated from high school. It's just too many years since high school! And a few years ago I started to lose height. (Sort of like a water balloon when you squish it down.) I've been shorter and fatter for longer than I thought.
Gert Loveday said,
June 14, 2014 @ 7:38 pm
Aha. Language is just the tip of the iceberg of thought, as demonstrated by the fact that you could think this sentence even while disapproving of it linguistically. Comparatives are a wonderful example of this. They don't operate on the horizontal linguistic string of fat-fatter-fattest but on a vertical plummet into your 'being in the world'. So 'fatter' refers to the whole complex of beliefs and reactions and experiences around being fat that you have, and 'longer' refers to your personal time line. So who cares about the syntactic string? superlatives perhaps put a lovely dome above the whole world, at least from the point where you're standing.
Rubrick said,
June 17, 2014 @ 2:05 am
You should definitely try to supress this kind of linguistically awkward thought. Fattery will get you nowhere.
He said, she said,
June 17, 2014 @ 9:20 am
My brother and I were leafing through a family photo album, commenting on how time has altered our physiques. Decades ago I was more slender than he but at some point my weight and flabbiness surpassed his. I had thought this was a recent development, but the chronological photo album revealed otherwise:
I've been fatter for longer than I thought.
In this context the sentence is unassailable, no?