« previous post | next post »

Yesterday, most of the comments on The communicative properties of footwear dealt with the gender associations of the word cute. This linguistic stereotype is often used as the basis of comic-strip humor, frequently in the context of shopping, as in this Foxtrot strip from a few years ago:

And (with a twist) in this Preteena from 6/24/2009:

But in fact, the word cute really is used much more often by women than by men, in modern American culture.

In a study based on a sample taken in 2004 (and described in "What men and women blog about", 7/8/2007), cute was the tenth most feminine word (as quantified by information gain), after hubby, husband,  adorable, skirt, boyfriend, mommy, yummy, kisses, and gosh. But male bloggers in that sample still used cute with a frequency of 83 per million words — it's just that the female bloggers used it with a frequency of 232 per million words — about 2.8 times more often.

In a large corpus of American English telephone conversations, the apparent femininity of cute was somewhat greater, at least as measured by the ratio of female use to male use — it was used 974 times in 15,685 female conversational sides, and only 214 times in 12,589 male conversational sides, for a ratio (corrected to account for the different numbers of conversational participants) of about 3.65. In comparison, the corrected ratio for shopping was only 3.01, and pink was only 2.12. (These transcribed conversations were mostly collected in 2003, but a significant minority were collected in 1990-91.)

In fact, cute was the most female-associated adjective among several candidates that I tried:

Female Male Ratio Corrected ratio
cute 974 214 4.55 3.65
adorable 57 13 4.38 3.52
gorgeous 241 76 3.17 2.55
lovely 168 57 2.94 2.37
pink 74 28 2.74 2.12
beautiful 1378 865 1.59 1.28
little 21017 15490 1.36 1.09
strange 1176 921 1.28 1.02
big 11500 9761 1.18 0.95
weird 1451 1731 0.84 0.67
cool 4951 6207 0.80 0.64
tough 1307 1765 0.74 0.59
lame 40 67 0.60 0.48

(The raw ratios of counts are multiplied by 12589/15685 in the column labelled "corrected ratio", in order to get the ratio of rates per conversation. The average number of words per conversational side was nearly the same for the sexes, with men being about 6% talkier, so to get per-word ratios, another small correction would be needed, which would raise the corrected ratio for cute to about 3.87.)

The gender divergence was greater for several nouns, with husband being more than 15 times commoner in the women's speech, and wife about 5 times commoner in men's speech:

Female Male Ratio Corrected ratio
husband 9168 484 18.94 15.20
boyfriend 1080 129 8.37 6.72
babies 570 122 4.67 3.75
shopping 1140 304 3.75 3.01
clothes 731 309 2.37 1.90
dinner 1093 507 2.16 1.73
shoes 608 382 1.59 1.28
baseball 1691 1720 0.98 0.79
dollars 6788 7880 0.86 0.69
cars 791 972 0.81 0.65
beer 230 388 0.59 0.48
girlfriend 612 1044 0.59 0.48
man 2889 5204 0.56 0.45
beers 25 75 0.33 0.27
wife 925 3786 0.24 0.20

[Update — Kenny Easwaran asks

[W]as there any interesting difference in use of these words based on the gender of the other conversational participant? It seems plausible that "cute" may be used more often with a female conversational partner than with a male one.

The answer:

female interlocutor male interlocutor
female speaker
762 in 10784
(7.06 per 100)
212 in 4901
(4.33 per 100)
male speaker 103 in 4901
(2.10 per 100)
111 in 7688
(1.44 per 100)

In other words, when a female speaker had a female conversational partner, there 762 instances of cute in 10,784 conversations, for a rate of 7.06 cutes per 100 conversations. And so on…

So indeed, both women and men tended to use "cute" somewhat more often when talking with a woman than when talking with with a man.]


  1. Kenny Easwaran said,

    October 24, 2009 @ 7:53 am

    Interesting that "boyfriend" and "husband" are so much more marked than "girlfriend" and "wife".

    [(myl) This is a symptom of a more general regularity. Though I can't prove it from this sample, I believe that in the case of words as in the case of clothes, it tends to be female-associated things that are marked. As a result, gender is signaled these days more by excluding men than by excluding women. That is, there are more words that are more strongly female-associated than there are words that are strongly male-associated, just as there are more kinds of clothing that are limited to women (or to men "dressing like women") than kinds of clothing that women can't wear (at least in some contexts).

    The analogy is far from exact — it's not at all the same sort of violation for a man to use the word cute as it would be to wear a little black cocktail dress and high heels — but I think there's some connection.]

    Also, was there any interesting difference in use of these words based on the gender of the other conversational participant? It seems plausible that "cute" may be used more often with a female conversational partner than with a male one.

    [(myl) Yes — see the Update above.]

  2. David said,

    October 24, 2009 @ 8:28 am

    I'm not sure this is specific to American culture. I don't have any data as firm as yours to back it up, but my hunch is that the Swedish word "söt" (literally "sweet", but corresponding closely in use to English "cute") is also much more in use by female bloggers than by male ones. I'm not sure how much evidence this is, but Googling Swedish-language sites, we can see that "söt" is used in 66% (439k) of the cases with a male subject ("he is cute") compared with 34% (224k) for "snygg" ("good-looking"), whereas with a female subject, the proportion for "söt" was merely 43% (416k vs. 541k for "snygg"). Of course there is some margin for error, given that babies can be cute irrespective of sex, and not all favourable comments on someone's looks are made by a member of the opposite sex. It would be interesting to know to what extent other languages have similar terms and whether their use reflects a cross-cultural phenomenon.

    [Search terms were on the form "han är söt" = "he is cute" with "är" and the more informal "e" both used in all cases.]

  3. Kerim Friedman said,

    October 24, 2009 @ 8:57 am

    Here is a paper on the "Power of Cuteness" in Taiwan:

  4. Aaron Davies said,

    October 24, 2009 @ 10:13 am

    @Kerim Friedman: interesting. I haven't read beyond the first page yet, but it already occurs me to to ask, is "ke’ai" cognate to "kawaii," is one a borrowing from the other, or are they only coincidentally similar?

  5. Kerim Friedman said,

    October 24, 2009 @ 10:36 am

    Aaron, I don't know, but I always assumed that "ke'ai" was in use before the "kawaii" fad, yet gained new life as a result of it.

  6. Kellen Parker said,

    October 24, 2009 @ 11:53 am

    I'm pretty sure that's the case, re ke'ai predating kawaii, though to be honest I have little to back it up beyond memories of hearing someone tell me as such.

    Either way it's beyond coincidentally similar. 可爱 is ke'ai and 可愛い is kawaii. 可爱 is literally "able to be loved" and without really knowing too much about Japanese, I can't imagine い does much to change that meaning.

  7. roscivs said,

    October 24, 2009 @ 1:03 pm

    The Japanese い is just a morphological suffix for adjectives (occasionally omitted when the adjective is uninflected). For most purposes you can consider 可愛い and 可愛 to be the same word in Japanese.

  8. David Eddyshaw said,

    October 24, 2009 @ 1:18 pm

    Japanese 可愛い (kawaii) originated as young-person slang for 可愛らしい "kawairashii".

    The first bit is the Sino-Japanese equivalent of the Mandarin ke'ai, so they are "cognate" in the rather peculiar way that Sino-Japanese is related to Chinese.

    I don't know about the history of this particular word, but certainly there has been borrowing of Sino-Japanese coinages made in Japan into modern Chinese, capitalising on the common script.

    Mind you, this looks like a pretty natural formation in Chinese anyway. But I don't suppose it would be surprising if a preexisting Chinese word picked up the associations of "kawaii" in the demographics in question ….

  9. Jerry Friedman said,

    October 24, 2009 @ 2:06 pm

    When I was teaching in the summer of '08, I was talking to a fifteen-year-old boy outside class, and I referred to someone or something (maybe an animal) as cute. He said "cute" wasn't a guy word. I asked, "Don't you talk about cute girls?" He said, savoring the phrase, "I talk about fine-ass chicks!"

  10. empty said,

    October 24, 2009 @ 2:13 pm

    I once wrote a letter to the editor on a serious subject, and my (heterosexual, male) grad student called it "cute". I don't get no respect.

  11. D.O. said,

    October 24, 2009 @ 3:31 pm

    Prof. Liberman, do I get it right that you think women tend to show off their "womaness" in speech more than men tend to show their "manliness". At least as far as word choice is concerned?

    [(myl) No — I'm not saying anything about that either way. Rather, I'm suggesting that if we were to do a systematic survey (which I haven't done) of how often the common words of English are used by one sex or the other, and sort the result by (say) the log of the female-use/male-use ratio, we'd find that there would be more words further from 0 on the female end of the scale than on the male end of the scale.

    There are lots of ways to test this in detail — different ways to quantify gender association, choices as to whether to count types or tokens, how (and if) to control for differences in topic choice, how to weight the distribution over individuals, etc. And I might be wrong in terms of any or all of them.

    But true or false, the claim is about words, not about women. If some version of the claim is true, it would be consistent with a variety of distributions of degree of gender-display on the part of women vs. men, since it would basically just say that women who choose to signal their sex verbally have a wider range of more effective options to choose from. ]

  12. D.O. said,

    October 24, 2009 @ 4:32 pm

    But if you hypothesis is correct, i.e. if there are more words very closely associated with females than words very closely associated with males than I do not see how their use might not be a gender signaling.

    [(myl) Those words would signal gender; but it's a separate question which women would use them how often.]

    I do not mean it as a psychological hypothesis (that is that women want to insert a reminder that they are women more often) rather than as an empirical observation. To put it blantly, if you hear someone saying cute, it's a woman, but if you hear someone saying lame you are not so sure.

    [(myl) Not exactly true: on the basis of the numbers I cited, one use of cute makes the odds about 3.6-to-1 that the speaker is female; one use of lame makes the odds about 2-to-1 that the speaker is male. That's a difference, but not a categorical one.]

    On a different matter. Why didn't you check the frequency of shoe/shoes? This was a conversation starter after all. I know, I should have done it myself and report the results, but ah, being a lazy ignoramus, it is too high a barrier for me to do it.

    [(myl) The numbers for shoes are in the second table, in the body of the post.]

  13. fiddler said,

    October 24, 2009 @ 6:39 pm

    I know three women who in addition to being attractive, bright, and fun, happen to be short and petite.

    They hate the word "cute."

    [(myl) How do they feel about the other adjectives that Preteena is taking refuge in the shower from?]

  14. empty said,

    October 24, 2009 @ 7:37 pm

    How do they feel about "petite"?

  15. fiddler said,

    October 24, 2009 @ 7:39 pm

    [(myl) How do they feel about the other adjectives that Preteena is taking refuge in the shower from?]

    Adorable, darling, sweet, spiffy, charming — I don't think these words have ever come up in this context. I'll have to ask. I can see adorable, darling, and sweet giving them the same heebie-jeebies as "cute," but probably not "spiffy" or "charming." (In fact, I don't think spiffy or charming really belong in that list.)

  16. David Deterding said,

    October 24, 2009 @ 7:48 pm

    The figures for adjectives seem to suggest that women use more adjectives than men. There are five adjectives used over twice as often by women (cute, adorable, lovely, gorgeous, pink) while only one adjective is used twice as often by men (lame). In contrast, for nouns, there is no such imbalance, as there are four female-preferred nouns using the double-occurrence criterion (husband, boyfriend, babies, shopping) and five male-preferred nouns (wife, beers, man, girlfriend, beer).

    Is this correct? Is this evidence to support a claim that women use more adjectives than men?

    [(myl) It's an interesting question, but the table in this post certainly doesn't support an answer. Such evidence as there is in the literature, e.g. this table from Thomson et al. "Where Is the Gender in Gendered Language?", Psychological Science 12(2) 2001, suggests that in fact a higher frequency of adjective-use is male-associated (here by being used more with male than with female interlocutors:

    Likewise in Sameer Singh, "A Pilot Study on Gender Differences in Conversational Speech on Lexical Richness Measures", Literary and Linguistic Computing 2001, male speakers used adjectives at a slightly higher rate than female speakers, though the effect size is very small:

    I'll see what I can do about POS-tagging the conversational transcripts used to make the counts in the body of the post, which would be a much bigger and more diverse sample (28,274 conversational sides rather than Thomson et al.'s 22 undergrads, or Singh's 30).]

    I know the frequency of occurrence of adjectives has been discussed in recent threads, and a claimed low frequency of adjectives in scientific writing has been shown not to be true. But a superficial analysis of these figures seems to support a gender-based imbalance. Is this accurate?

    [(myl) Probably not.]

  17. Alan Gunn said,

    October 24, 2009 @ 8:01 pm

    Maybe it's just us lawyers, but most of the time I hear "cute" it's a sort of synonym for "clever," but without any suggestion of admiration. Successfully using inappropriate literalism or maybe taking advantage of someone's ignorance would be examples of the sort of thing that could be called "cute." I can't imagine using it any other way myself, save perhaps in connection with very young children or cuddly animals.

  18. David Deterding said,

    October 24, 2009 @ 10:04 pm

    Table 1 from the Singh (2001) paper does indeed show there is little difference between men and women in overall adjective usage. But what it also seems to suggest is that there is far greater variation among men (SD=1.66) than women (SD=0.77) for adjective usage. It's hard for me to interpret this — are some men heavy adjective users while other men use adjectives more sparingly, but women vary less in this respect? And if this is accurate, why should such an imbalance occur?

    [(myl) You can't conclude much about men and women in general from about 1,000 words of conversation from each of 13 male and 17 female subjects. In particular, I wouldn't count on the standard deviations of proportions of adjectives generalizing.]

  19. Rubrick said,

    October 25, 2009 @ 3:36 am

    I'd bet the female/male "cute"-usage frequency imbalance would be even greater if we excluded uses where the adjective was directly describing a person of the opposite sex. In my (American) experience it's perfectly "male" for a man to refer to a woman as cute, meaning simply attractive; somewhat less so to describe her hat as cute; still less to describe his own hat as cute.

    Somewhat tangentially, It seems as though during my lifetime there's been a transition of "cute" when describing an adult from meaning "adorable", with overtones of "childlike, precious", to something closer to simply "hot". No idea if this is the Recency Illusion at work, or perhaps even a regionalism (I moved from the east coast to the west in my twenties).

  20. J. Goard said,

    October 25, 2009 @ 5:07 am

    For me, the most irksomely non-guy word in the first comic, by far, is top 'upper-body clothing'. It's not going to show up in a simple search, of course.

    I don't think I'd be able to control my laughter if I heard a straight male acquaintance use top instead of T-shirt, blouse, etc.

  21. Graeme said,

    October 25, 2009 @ 6:33 am

    Hypothesis: correlate hours invested in 'shopping', 'clothes' and 'babies'/'children' and you'll explain the tendency to verbal cutification. I say that with the greatest respect as a dad who went part-time from academia to raise two girls.

    (This is not to decry 'cute'; on the contrary it has stronger/clearer associations and connotations than other positive, if more euphonious adjectives such as 'beautiful' and especially 'sweet', 'gorgeous' and 'lovely'. If anything it is overused beyond its usefulness. Generally the problem is a blanding and blending of adjectives, into a 'nice' puddle, when there are so many more tailored terms like 'elegant', 'delicate' etc – not to mention effective genetics like 'fetching' – falling by the wayside.

    If this sounds elitist, put it this way. Do advertisers often employ adjectives like 'cute', let alone 'nice'?)

  22. Nick Lamb said,

    October 25, 2009 @ 6:52 am

    J. well I am male, and can easily imagine using that word, and the reason would be a failure of ontology. I'd begin a sentence intending to complement an article of clothing I liked, "Oh hey Sue, that's a nice …" and somewhere around the middle of that sentence I'd start trying to pick a noun – hmm, it's not exactly a blouse, and why does it fasten at the side like that… and with these deliberations unfinished as the need for the final noun becomes pressing, I'd choose the safe but boring 'top'.

    I'd have assumed that 'top' would be the most common choice for men because it saves them needing to have an ontology of female clothing, and that would fit with a general level of ignorance about women's clothes that is stereotypically expected. But I don't have proof of that, and so I defer to J's observations pending any real data.

  23. Nick Lamb said,

    October 25, 2009 @ 11:44 am

    Ugh, s/complement/compliment/

  24. John Cowan said,

    October 25, 2009 @ 9:04 pm

    Alan Gunn: I suspect that that, like much lawyerspeak, is a survival from an earlier stage of English. In the 19th century, cute was short for acute (and often spelled 'cute) and did primarily mean 'clever'. Dickens makes a blacksmith in Barnaby Rudge say "He will be a 'cute man yet", and in Mark Twain's Pudd'nhead Wilson, whose protagonist is a lawyer, we find this (emphasis added):

    For some years Wilson had been privately at work on a whimsical almanac, for his amusement—a calendar, with a little dab of ostensible philosophy, usually in ironical form, appended to each date; and the judge thought that these quips and fancies of Wilson's were neatly turned and cute; so he carried a handful of them around one day, and read them to some of the chief citizens.

    The OED points out the similarity to cunning, which likewise originally meant only 'clever' but also added the sense 'charming', though more usually applied to things than people.

  25. jo said,

    October 26, 2009 @ 7:47 am

    Just a few words about 可爱 (ke'ai) and Japanese kawai-i (sometime written 可愛い). Kojien (reasonably authoritative Japanese dictionary) unambiguously states that 可愛い is an ateji (i.e. characters assigned to a pre-existing word without regard to its etymology). Given the existence of Chinese 可爱, it seems unlikely this particular ateji is pure chance, but neither is it any mark of etymological relatedness. Wikipedia (same .ja link as below, if it shows up, to the page "可愛い") gives an etymology involving kao (face); this, if correct, rules out potentiality (可) and love (愛) in the evolution of this word.

    First aside: the 可愛い ateji are neatly mirrored by others such as 可笑しい (lit. laughable) for okashii (strange).

    Second aside: that kawaii developed from kawairashii (even if their relative frequency of use might have changed from the former in favour of the latter – who knows?) seems implausible at best; rashi(i) is a productive affix in its own right. Kojien and wikipedia (yes, I know, not too trustworthy) do not give any indication that this is the case.

  26. Cath the Canberra Cook said,

    October 26, 2009 @ 8:23 am

    The old meaning of cute as clever still gets some use in Australia. A "cute trick" is a clever and non-obvious action, something a little unexpected that works nicely. A clever workaround. The first one I googled up (with the .au restriction) is an example of a literal trick –
    The second one that emerges from the sea of "cute trick or treat x" is this familiar usage (to me as an Australian former maths nerd) "By a cute trick the cosine interpolation reverts to linear if applied independently to each coordinate."

  27. Alexandra said,

    October 26, 2009 @ 11:26 am

    @ Nick Lamb: I would just say "shirt."

    I am female and American, but would never actually say "top." To me, it feels like either advertising-speak or something out of the movie Clueless.

  28. Barbara Phillips Long said,

    October 27, 2009 @ 10:41 am

    Alexandra — I'm an American female and I use "top" all the time, because I don't consider the tops I wear to be t-shirts, but they aren't blouses or sweaters, either. They're mostly knit tops, but I have never called them tees because they don't have the traditional crew neck finished with a ribbed neckband. A few are woven and aren't knits.

    Apparently blouse, for me, means something with buttons in either front or back, shirt is something tailored with buttons in the front, t-shirt has the neckband, polos have the knit collar and placket, and knit or woven things that go over my head and cover the top half of my torso are, well, tops.
    = = = = = = =

    Unfortunate use of "cute" by male:

    This is from a recent Carolyn Hax chat (she's the advice columnist for the Washington Post) and the remarks of other people as they chime in:

    Duluth, Minn.: My wife is afraid of everything … flying, confined places, someone breaking into our house and many other things. I use to think it was kind of cute but with children I worry that it is impacting them. Moreover, I worry that she is not allowing them to grow because she is overprotective. They can't do many of the things (e.g., rides bikes, walk to friends houses, or play basketball on a quiet street with construction cones set up as roadblocks) that the other kids in the neighborhood do. I don't think we can live in fear of miniscule risks but I get nowhere when broaching this with her. I am mostly concerned about my oldest who is almost 11 (the youngest is 7). Any thoughts?

    Carolyn Hax: "Cute"? To be living, at best, half a life, and all of it in fear? …

    (omitting chat sections) …

    To Duluth Again: : And stop being so PASSIVE. You worry that SHE's not letting them grow up? Where have you been this whole time while she was laying out unreasonable rules?

    And on what planet was it cute that your wife lives in terror?

    Or did her exaggerated fears just make you feel more secure and manly?

    Carolyn Hax: I'm guessing there was a lot of (false) security sought in role-playing. Home/children/fainting couch were her turf, and protection/breadwinning were his. Terrible trap for all involved, but especially the kids. …

    (omitting several exchanges by Duluth and others about his situation) …

    NYC, N.Y.: re: Duluth

    There's a harshness to the responses posted from the peanut gallery which makes me wonder why you chose those particular responses to post. This is someone who is coming to you for advice about his dawning awareness that the love of his life is not a fit parent. This has got to be a difficult place for the non-primary caregiver to be. You gave him, among other things, the off-topic accusation that he didn't come to his worries sooner because his partner's worry made him feel manly.

    Though I think you have good points today for him, I also think you threw him to the wolves a bit, and I don't blame him for his defensiveness. Please remember: people are coming to you for help because they don't know who else to ask.

    Carolyn Hax: Thanks. Sometimes I don't see this forest while I'm up to my ears in trees.

    The frenzy was off his calling his wife's fears "cute," which warranted mention but not a mob with torches and pitchforks.

    To answer your question, I choose responses based on what appears to be a group tone–the reasoning being, if this is what a lot of people are saying, then it represents something I missed/underplayed. It's not intended to whoop people up, but instead to round out my answers.

    I will keep your concerns in mind as I choose, thanks again.
    = = = = = = =

    The full chat is at:

  29. Carl Offner said,

    October 28, 2009 @ 9:21 pm

    From the book Teacher by Mark Edmundson — a phenomenal and (in my opinion as a former classroom teacher) overlooked book:

    "I had come late to class that day, so I was compelled to sit on the outside of the circle. I slid my seat up behind two of the girls (I can't recall for certain which ones — Nora and Carolyn, maybe), who seemed not to know, or more likely not to care, that I was close by. Their subject for the day, addressed in the subtlest whispers, was who was and who was not cute. The conversation was a game, played a little like tennis. One of the girls would serve up a boy's name, and the other would acquiesce and say that yes indeed, he was cute — that is, attractive, sexy, desirable (1950s euphemisms were still in play, though, so *cute* was the word), and thus the server would pick up a point. If the other girl blanched and cringed or offered an expression proximate to vomiting, then the point was lost."

RSS feed for comments on this post