This is a question about pairs of anagrams capable of flavoring puzzles as memorable and surprising because the two words are so different.

One of my favorite such pairs is chesty and scythe. These sound and look very different:

  • two syllables; one syllable

  • c and h together to represent the sound /tʃ/, with t representing /t/; t and h together to represent /ð/, with c immediately before s to represent /s/ (or, if you prefer, silent)

  • e pronounced; e silent

  • y pronounced two different ways

  • y as suffix and in the root

  • one pair of consecutive consonants (with that term understood as it should be, to denote a non-vocal phoneme); no pairs

  • two monophthongs, no diphthong; no monophthongs, one diphthong.

The only letter pronounced the same way in both words is s, and even then you could reasonably say that the letter group representing the phoneme /s/ in scythe is sc rather than just s.

They are also a long way apart semantically.

The pair is practically a work of art!

The pair bedroom and boredom is quite good, but hardly in the same league. Conversation and conservation is at the other end of the scale.

What other pairs of very different sounding and looking words are anagrams of each other? What nice metric might we use for difference?

And what about triples or larger sets?

(I was caused to think of this question after tackling @BmyGuest's challenge question. Add r to toenail and rearrange to get relation; t to senatorial, alterations; or in @Psybin's answer, i to castle, elastic.)

Apart from {chesty, scythe} and {admirer, married} (found by @Quark), are there any other pairs of single-word anagrams, with say five or more letters, in which no consecutive letter pair appears in both words? Four-letter examples include part and trap, too symmetric to get a high score for the difference.

  • $\begingroup$ This isn't intended as primarily a puzzle, more a request for help in getting to a good definition of what kind of difference (not semantic linkage) makes anagrams memorable and surprising, which of course requires lots of nice examples. $\endgroup$ – h34 Jan 25 '15 at 11:53
  • 3
    $\begingroup$ My is "Banach-Tarski" and its anagram, "Banach-Tarski Banach-Tarski". $\endgroup$ – KSmarts Jan 26 '15 at 17:16
  • $\begingroup$ Possibly of interest: scoring of all anagrams according to the number of pairs of non-matching consecutive letters: blog.plover.com/lang/anagram-scoring.html $\endgroup$ – Mark Dominus Apr 19 '17 at 20:36
  • $\begingroup$ Thanks, @MJD ! There are some excellent pairs there, including cinematographer-megachiropteran, dispensable-piebaldness, coprophagist-topographics, masculine-calumnies, and earringed-grenadier. $\endgroup$ – h34 Jun 27 '17 at 12:33
  • $\begingroup$ sectional coastline $\endgroup$ – Daniel Mathias May 14 at 21:19

I'm not a linguist so I couldn't tell you about diphthongs or phonemes, but here are a few:

generate --> teenager

marginal --> alarming

interests --> triteness

infringe --> refining

bargained --> gabardine


Here's a good one I found:

altering --> relating --> alerting --> triangle --> integral

  • $\begingroup$ These are all pretty nice, but in each pair except the last the two words share a vowel sound (which can be either the first or second in interests - assuming you pronounce it as a trisyllable - depending on how you pronounce triteness) and a string of 2-3 letters, albeit in the 3rd and 5th pairs not pronounced the same. Which takes us some way towards a possible metric. $\endgroup$ – h34 Jan 25 '15 at 11:36
  • $\begingroup$ A phoneme is a sound unit capable of bringing about a change of meaning, so for example horse has 3, all different, and lull has 3, the first and third counting as the same even if they are pronounced noticeably differently from each other. A monophthong is a vowel sound you can hold (try it with cart); a diphthong you can't hold because it's made up of two vowels, one gliding into the other (try holding the sound between the two consonants in game, tone, now or kite.) $\endgroup$ – h34 Jan 25 '15 at 11:47
  • $\begingroup$ Note that after the question was edited this limitation would eliminate some of your examples: "in which no consecutive letter pair appears in both words". I thought of "astronomer" and "moonstarer" which only works if you accept my claim that I swapped both o's and both r's! $\endgroup$ – AdamV May 27 at 12:51

Ross Eckler's Long Well-Mixed Transposals (Word Ways, Feb 1976, p.20) addressed this problem. The examples in this answer are taken from Eckler's article. Here are the longest word-anagrams whose words do not have a single bigram in common. The words are 14 letters long.

nitromagnesite - regimentations (Murray Pearce, Enigma, Aug 1971)

rotundifoliate - titanofluoride

The latter was found by Dennis Ritchie in 1973. Eckler does not mention it as having been published, so perhaps Eckler's article is its first publication. Some near misses using 15-letter words:

dechlorinations - ornithoscelidan (Hercules McPherrin, Enigma, Jun 1925)

cinematographer - megachiropteran (Hercules McPherrin, Enigma, Dec 1927)

The latter two 15-letter words have the bigrams er and ra in common, but in megachiropteran they overlap, so you'd still need to cut each word into 14 bits if you wanted to turn one into the other by permuting the bits.

"Transposal" is Eckler's term for anagrams of one word to one word, without regard to the words' meaning, as opposed to anagrams which have some sort of semantic link, e.g. H.M.S. Pinafore $\rightarrow$ name for ship, or angered $\rightarrow$ enraged.


medical and decimal are pretty close, but claimed and declaim are also anagrams of them. Pick any pair. In particular, claimed/decimal and medical/declaim both satisfy the "no pair of letters in the same order" criteria.

I may come up with some more later...


Many of the anagram pairs in this hall of fame:

Dictionary = Indicatory

Elvis = Lives

Listen = Silent

Admirer = Married

Although the letters don't really have different sounds.



  • $\begingroup$ Answers that are just links aren't exactly that good, if the links break than this answer is no longer useful. Would it be too much if I asked you to add in some of the relevant information of the link in, then use the link as a reference/source? $\endgroup$ – warspyking Jan 25 '15 at 4:20
  • $\begingroup$ I'll undelete my post since warspyking made the effort to edit it with stuff from the link. $\endgroup$ – Quark Jan 25 '15 at 5:38
  • 1
    $\begingroup$ I think it is a bit overdone now. Can you pick a best-of with respect of the asked question (plus the link)? I think any answer that long must have a very justifiable reason. A lot of the anagrams from the site do not fit the "pairs of words" or dissimilarity property. $\endgroup$ – BmyGuest Jan 25 '15 at 7:37
  • $\begingroup$ @Quark I was only trying to e courage you to fix it up... $\endgroup$ – warspyking Jan 25 '15 at 10:01
  • 1
    $\begingroup$ Thanks for this, but the question is about anagrams of single words. I'll edit to put those pairs at the top. Admirer and married make a good pair - they don't share a string of 2 letters and they don't share a vowel sound if you pronounce the a in admirer as /ə/ (schwa.) But they are (at least a lot of the time!) in the same semantic field. In fact a lot (all?) of these pairs have some semantic closeness. $\endgroup$ – h34 Jan 25 '15 at 11:38

Would this pair be an interesting addition:


3 syllables vs 2, the C and H, and the A pronounced very differently.

But maybe they are too close in meaning - after all, they are both led by a man with a stick!


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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