From Ben Zimmer comes news of a neat offshoot of his Visual Thesaurus:

When we launched the Visual Thesaurus Spelling Bee this past summer, we knew there was a built-in interest, but the response was still surprising. So far there have been 15,000 players who have tried their hand at spelling a grand total of 500,000 words. It’s clearly habit-forming, with many repeat visitors. The reason why it’s so addictive is that it’s been designed to be adaptive, so the more words that are spelled correctly, the more difficult the words become. And conversely, if you’re not a great speller, the words will get easier and easier. That way a player will always be quizzed at the appropriate skill level — from the orthographically challenged to the most expert spellers.
As more and more players try the Bee, the game has steadily improved based on data collected on how words are spelled. Words are being continuously reanalyzed for difficulty based on how spellers fare. Every five minutes, words are rescored for difficulty taking into account the latest data from the Bee spellers. That means there’s an increasingly better fit to different skill levels. …
For each word, a graph is generated to plot the distribution of right and wrong answers across different skill levels. Then a curve is drawn to fit the data. If that curve rises very steeply, then the word is a good “discriminator”: it’s an accurate way to separate the good spellers from the bad spellers.

I’ll give you a word of warning so you don’t stumble the way I did: be sure to read the definition on the lower right before trying to spell it.


  1. Verdammte Yankees and their schwas! I was cruising along until I got sporophore. No matter how many times I played it, I heard it as sporəphore. I had thought I was so clever when I correctly parsed the word that I heard as “careless” as “garrulous”, but the woman’s refusal to enunciate that “o” was my undoing.

  2. So many of those words are words that I knew from writing, but had never known how to pronounce (even if I thought I had!). In some cases, I was able to make the connection anyway; in other cases, not. “Brougham”, in particular, was nothing like I’d imagined.
    Neat indeed!

  3. mollymooly says:

    The audio for “bumptiously” was unclear and the definition was “in a _____ manner”. Harrumph!

  4. mollymooly says:

    And I’ve just guessed “brookline” for “brooklime”. Are they hosting a spelling bee or just chumpsourcing their quality control?

  5. An Australian accent would have been helpful, at least for this Australian!

  6. Why “bee”? (I’m sure I could look it up.)

  7. You could, but you wouldn’t get an answer (at least, not a trustworthy one); the origin is unclear. Here is one guess, but until someone comes up with proof that such competitions were actually called “boons” or “beens,” it’s only a guess.

  8. marie-lucie says:

    “bee” ‘a social gathering for the purpose of a common activity’
    The link found by LH conflates several sources, not all of them reliable. I am not an expert on Middle English, but if there was indeed a word “bene” it cannot be the source of “benefit” (derived from Latin). “boon” meaning “favour” is possibly from French “bon” (which, besides ‘good’, can mean ‘voucher’ or ‘bond’), but “boon” in the meaning “bee” might have a different, non-French source, closer to that of “bene” (ME sources are not all from the same dialect, hence the same word often occurs under different forms. “boon” would not have morphed into “been” in crossing the Atlantic, but both forms might have come over with different groups of immigrants).
    The other quoted alternative seems more likely, and it comes (although misquoted) from the list of Indo-European roots:
    (root *bheug) 1a. bee2, from Old English be:ag, a ring, a root also found in b. bagel, from Old High German boug, a ring.
    (here I use “:” instead of a macron [line over the letter] to indicate length since the letter e with macron does not reproduce in copying the word and I don’t know how to do it; LH’s source did not pay attention to this and quotes OE as “bag” instead of be:ag).
    It seems likely then that bee originally referred to a circle of persons, meeting in order to ease the boredom of a repetitive activity (as with spinning or husking bees). The form “bene”, later “been” (if related) could have been a plural form (with n as in “oxen”).
    These are my opinions, but they should be checked further, for instance in a Middle English dictionary.

  9. michael farris says:

    I never was (and never will be) a real star speller. I went through 50 words and got a 700 score (could be worse, could be better). If I didn’t get it in three I just gave up (figuring that doubling and shifting letters was a waste of time).
    Pronunciation gripes: the first vowel in the recording of laminectomy sounds (to me) like it rhymes with hem and not ham (which I would have expected). There also appeared to be an intrusive l in another word (I forget which but it was before an m)
    My weirdest miss was ‘remunerative’ which I’ve always thought was ‘renumerative’ (quick superficial googling shows I was not alone in my ignorance, which is a comfort).

  10. The pronounciation is a problem, though it may in part be the poor speakers on my laptop. But I definitely had trouble with the female voice, with “dahlia” and curiously “monte” (card game) which sounded like “mantee” to me. I just couldnt make sense of the audio. But it’s rather fun, none the less.

  11. Hmph! When I reached a plateau at 800, with a maximum straight run of 21 right, I did not press on. Complaints:
    1. The pronunciations were just not good enough, even on my perfectly adequate system and with negligible ambient noise. To add to gripes vented above: I couldn’t hear a difference between opening ”per-” and schwa-like ”a-”; ”extra” (surely a trivial component to identify normally) sounded like ”exo”; and so on.
    2. The written clues were uneven and quirky in their logic, and sometimes downright misleading.
    3. Even allowing for one’s individual customised trajectory through the battery of questions, there are not enough words of intermediate difficulty for the test to make reliable discriminations of ability in the mid-range. It’s easy for a tester to include genuine oddities from Middle Algonquian, but hard to approach them gradatim.
    Combined, these flaws ensure that the instrument falls short for testing at and above high-middle competence.
    I regard myself as an excellent speller. This bee gets a C-minus from me. Still, I agree it’s a bit of fun.

  12. Red-faced, I draw back some of my gall now that I read properly and find 800 to be the maximum possible score. :)
    I refine my diagnose of the difficulties with intermediate-level words. I suppose a bank of 500,000 must be enough! The test seems rather not to bother with fine discriminations in the high band.

  13. Ahem… “refine my diagnosis“. And retract again. It doesn’t seem to be a bank of 500,000 words! This is seriously misleading:
    So far there have been 15,000 players who have tried their hand at spelling a grand total of 500,000 words.
    No, most likely there were 500,000 attempts at words (by fewer than 15,000 punters, as we are more or less told – though even that is not a safe conclusion); or perhaps 500,000 responses including multiple attempts at somewhat fewer presentations of words.

  14. Thanks for all the excellent feedback, everyone.
    Re pronunciations: As the Bee page explains, listening with headphones will greatly improve your ability to hear nuances of pronunciation. We’re also re-recording any words that people find to be unclear, so please do use the feedback link at the bottom of the page to report any problems.
    Re Noetica’s comment that “the test seems rather not to bother with fine discriminations in the high band”: We’re actually making finer and finer discriminations in the high band, rescoring those high words based on the data we get from excellent spellers such as Noetica who are able to get to the 800 level.

  15. I couldn’t concentrate on it. So why can concentrate on Free Rice, and even get lost in it?

  16. Sure Ben. Thanks for the test! We all love getting our teeth into these things, and the LH pack can get pretty ferocious.
    The pronunciation issues are quite hard to address, even when the problem of different varieties of English is accounted for.
    I do think the written clues need re-visiting, especially where they are frankly misleading – such as one with something like this form, with a simple blank at the end:
    An adjective for ________.
    ¿Qué? Tells us almost nothing, and certainly diverts us from what at first seemed like a plausible answer.
    Of course the test serves its promotional, educational, and entertainment purposes. But with words recycled like that, even within a single test session, you are confounding spelling ability with ability to recall, and with motivation, obsessiveness, and game-savviness. This must seriously compromise testing in the high range, so I fear that the clever re-weighting of words is not as effective as it might otherwise be. That re-weighting is valuable, though; it could be used to identify relative difficulty levels at which there are insufficient words. If it is not being used that way, why not consider such an extension? This could ultimately lead to finer discriminations at middle and high levels, and possibly enable revision of the upper limit in the scoring.

  17. One of the women has a really strange way of saying things. To me, it sounds like she begins “utile” with a B or an M. There were several other ones where it seemed like she got a consonant flat-out wrong.

  18. Ben: If you consider that you are addressing an international audience, how about having two pronuciation options – American, or British R.P. ? Double the trouble for you, but likely to be more accessible to many readers outside the U.S.
    I thought I could cope with the American accent, but in this context, it’s much more difficult that I generally find it. Witness my “monte”/”mante” problem – they are quite different in US and RP, where the o is as in Montana.

  19. All right, there are several words on here that I even know and have heard pronounced by other people and they don’t say them like the recordings on this site. Also there were times where I thought she said an [h] and she said a [p] or something. I fail at this.

  20. I think I’m giving up on “myrrh,” which I frustratedly wrote “neur” for because I had no idea what she was saying.
    Also some of the definitions are things like “having a lot of _____” or “being ____like” which is completely useless.

  21. I found a pronunciation mistake, at least according to Webster’s Third. The test said “fair-a-SIGH-a-kal” while Webster’s and I say “fair-a-SAY-a-kal.”

  22. Also “fort-MONT” versus “fort-a-MEN-tay.”

    In the American way, it is pronounced as only protrusal would be to the rest of us (who would rhyme the last syllable with aisle). Temporarily forgetting this difference, I missed protrusile even though I would certainly have got it on etymological grounds.
    Second syllable very strangely pronounced, and not as in SOED, OED, or M-W’s dictionaries.
    First syllable with /ei/ is not supported in SOED, OED, or M-W’s dictionaries.
    A fascinating convergence, here. I put foveolate, and was most surprised to have it rejected. In fact both words exist, and are from quite distinct Latin sources: foveola, diminutive of fovea “pit”; and faveolus, diminutive of favus “honeycomb”:
    (OED: “Marked with little depressions or pits; pitted.”)
    In OED, SOED, Webster’s 3rd International (W3I).
    (OED: “Honeycombed, cellular.”)
    In OED, W3I.
    Certainly the pronunciation given in the Visual Thesaurus bee warrants either; and the written clue is “pitted with cell-cavities (as a honeycomb)”, which surely fits either word.

  24. Noetica, that’s a truly remarkable convergence!

  25. marie-lucie says:

    … quite distinct Latin sources: foveola, diminutive of fovea “pit”; and faveolus, diminutive of favus “honeycomb”
    The convergence must date back to Latin, probably of a late period: foveolus from fovea is well-formed, on a stem fave-, but for a classical formation the diminutive of favus (stem fav-) should be favulus not faveolus. The convergence is understandable if a honeycomb is viewed as a mass of little pits.

  26. John Emerson says:

    From the point of view of English vocabulary, I’m trying to imagine how we could get farther lost in the weeds than we are right now.
    I mean that in a good way! Getting lost in the weeds is fun!
    The etymology of “sack” is the best I can think of. From Hebrew (story of Joseph?) but before that, maybe from Akkadian or Sumerian.

  27. John Emerson says:

    From the point of view of English vocabulary, I’m trying to imagine how we could get farther lost in the weeds than we are right now.
    I mean that in a good way! Getting lost in the weeds is fun!
    The etymology of “sack” is the best I can think of. From Hebrew (story of Joseph?) but before that, maybe from Akkadian or Sumerian.

  28. A.J.P. Crown says:

    Whereas ‘bag’ is Norwegian, or so I’m told (constantly).

  29. 670/ Got more at times but not for long :( Shame on me.

Speak Your Mind