Banana, alfalfa, entente: these are words containing an overlap, that is, the pattern XYXYX where X and Y are replaced by letters or sequences of letters. The idea is that the two occurrences of XYX overlap.

Is there any 5-letter English word like that? Any other examples besides the ones I mentioned?

  • It rather depends what you accept as a "word". You could have UrbanDictionary's bobob (acronym for Big Ole Bitch On a Bike), for example. But I think the question is Off Topic. Jul 19, 2014 at 15:29
  • 2
    @FumbleFingers "Off Topic", how so? Jul 19, 2014 at 15:56
  • 3
    Because it's not about English Usage. You're just asking for a list of words matching some arbitrary orthographic criteria (it's irrelevant that imho it would be an empty list unless you include "words" not normally found in dictionaries). Jul 19, 2014 at 16:02
  • 2
    I think I can safely assure you there's no inherent tendency for languages to "tolerate" such replications in the sense you imply. English definitely encourages certain "reduplication" forms. But your examples are just "quirks" - they say nothing about how English "works". Jul 19, 2014 at 16:53
  • 1
    @FumbleFingers: If you were to close, surely it should be because this is a "duplicate question" ;).
    – JohnLBevan
    Aug 20, 2015 at 11:24

1 Answer 1


In the FreeBSD operating system /usr/share/dict/words list, the following words of 6 letters or fewer match the pattern:


I believe the word list is based on (but not identical to) the 1934 Webster's New International Dictionary, Second Edition.

I generated the list with:

egrep '(.)(.)\1\2\1' /usr/share/dict/words | awk '{ print length, $0 }' | sort -n | cut -d" " -f2-
  • I guess "banana" is the only one on that list that's really well known. Jul 19, 2014 at 16:44
  • 3
    I'd definitely say it's the best-known, but rococo might meet the threshold of well-known, and Macaca had an unfortunate moment in the spotlight recently.
    – coneslayer
    Jul 19, 2014 at 16:48
  • 1
    @FumbleFingers Running (a derivative of) that pattern on the OED yields these additional six-letter solutions: berere, derere, desese, hadada, halala, Ibibio, mañana, marara, matata, ototoi, patata, pewewe, sememe, shshsh, tarara, wenene, zanana. As you see by mañana, I ignored diacritics (and case). The change compared with the original to make it insensitive to diacritics or case was this (?xi) (?=(\w)) \X (?=(\w)) \X \1 \pM* \2 \pM* \1 \pM* run on NFD data. BTW, seven-letter OED solutions are awaward, bararag, heleles, ngarara, ogogoro, ototomy, phememe, phenene, ukelele.
    – tchrist
    Jul 19, 2014 at 22:45
  • 3
    @FumbleFingers You appear to be lost. This is my answer to the question. Meta is down the hall, second door on the left.
    – coneslayer
    Jul 20, 2014 at 1:49
  • 2
    @FumbleFingers And if it were emigrated to SO, it would be closed for asking whether something was an English word. If it were sent to Linguistics, then it would be closed for asking about English-specific matters instead of more language-neutral ones, or perhaps about things deemed beneath the dignity and interest of that august community. I’m sure we can think of other SE sites where the question’s germanity could also reasonably apply — yet where it would be met with scorn equal to your own. Nobody has a home for these sorts of questions.
    – tchrist
    Jul 20, 2014 at 2:36

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.