I don't know how 'worthy' came to be pronounced with a /ði/ whereas 'healthy', 'wealthy', 'stealthy' etc., are pronounced with a /θi/. I am aware that theories of pronunciation in English are far too complex and often seem illogical. But with a curious student in my front wanting to know why 'worthy' has the /ði/, my task becomes very difficult.

I would appreciate if you could let me in on something that might help.

  • 1
    Merriam-Webster gives two pronunciations of earthy: one with /ði/ and the other with /θi/. I think it might have to do with the /r/. (The other dictionaries I checked just have /θi/, which is the way I've nearly always heard it.) Commented Jan 30, 2013 at 16:30
  • 2
    @PeterShor Swarthy would be another with /ði/, and I don't think there's many others that end in -rthy
    – Marcus_33
    Commented Jan 30, 2013 at 16:38
  • For those, like me, that are IPA deficient: ð is a voiced dental fricative and θ is a voiceless dental fricative. Commented Jan 30, 2013 at 19:06
  • 3
    Or for the people who don't want the phonetic names: ð is the beginning of "this" while θ begins "thin"
    – user10893
    Commented Jan 30, 2013 at 19:56
  • It doesn't help that the difference is acoustically subtle. Helped me to try worthy then worth-y. This question applies to swarthy.
    – deadly
    Commented Feb 1, 2013 at 16:08

2 Answers 2


This is a fascinating question, and I have no idea how that particular voicing came to be. Roger Lass is probably who I'd ask about that.

I was interested enough, though, to look at the words in Webster's 2nd ending in -lthy and -rthy.
( egrep rthy$ webster2.txt, etc. )
They turn out to have a very interesting pattern.

In each group (which are roughly the same size, in terms of number of root words), there is, first, a smattering of individual words like swarthy and filthy, including healthy and unhealthy, stealthy and unstealthy.

And then there comes a flurry of compound words made from a single root word; both of these words refer to the same concept, though in two flavors:

  1. wealth (always pronounced with /θ/, even in -wealthy /'wɛlθi/)
  2. worth (pronounced /θ/, but not in -worthy /'wərði/)

As can be seen, words ending in -worthy are vastly more diversified and specialized, to the extent that -worthy has become something of a libfix, like -ectomy. This may or may not have anything to do with the voicing; I wouldn't know.

The Data

. . . R . . . . . .. L . .
birthy . . . . . filthy
earthy . . . . healthy ~ unhealthy
forthy . . . . stealthy ~ unstealthy

airworthy . . . . . overwealthy
Allworthy . . . . . subwealthy
blameworthy . . superwealthy
bloodworthy . . . ultrawealthy
bribeworthy . . . unstealthy
evenworthy . . . unwealthy
faithworthy . . . wealthy

  • 2
    Some thoughts: 1. in OE, in a position a stressed vowel+fricative+(liquid)+unstressed vowel, fricatives were always voiced. 2. First documented occurrences (OED): worthy (c1220) - the oldest (!), wealthy (1430), earthy (1398), healthy (1552), filthy (1382) etc. Not sure whether this is relevant though.
    – Alex B.
    Commented Jan 31, 2013 at 0:15
  • 2
    Maybe, voiceless /th/ itself is anomalous in words like earthen, earthy, filthy, frothy, lengthen, stealthy, wealthy, strengthen. Many ModE words with final voiceless /th/ once had voiced /th/ followed by a vowel. After having lost the final vowel, voiced /th/ became voiceless: beneaðn > beneath, eorðe > earth, piða > eahtoða > eighth.
    – RainDoctor
    Commented Jan 31, 2013 at 4:46
  • Now that John Lawler mentioned libfix (and I heard it for the first time), are there any actual instances of lexical words, not affix types like '-ectomy', which have become libfixes and their pronunciation has changed, if only slightly, for that very fact?
    – user32480
    Commented Jan 31, 2013 at 4:49
  • 1
    /θ/ and /ð/ are notorious for having by far the smallest functional load of any pair of phonemic consonants in English. There are only two minimal pairs (thigh/thy and ether/either), and they're both flaky -- thy is archaic and either is pronounced with a different vowel by many speakers. Plus there's significant allomorphic variation (smooth/smoothe, loath/loathe, bath/baths) and significan free variation (with as /wɪθ/ or /wɪð/), and mostly they occur in lexical items with other significant properties, like Greek borrowings for /θ/ or closed classes for /ð/. Commented Jan 31, 2013 at 17:51
  • @RainDoctor: that seems like a possible lead. A related case of inconsistent use in Modern English of voiceless or voiced fricatives before the suffix -y is discussed in this question: Pronunciation: ‘lousy’ vs. ‘mousy’. Why?. The answers there suggest it is based on the time in history when the -y suffix was added to the word. Alex B.'s data seems to be quite relevant.
    – herisson
    Commented Nov 10, 2015 at 6:35

In the case of /'wɛlθi/ vs. /'wɚði/, we're looking at more than one hypothetical rule, and the more I look into it, the more it screams ?exception. Of course, much like OP says, English phonetic rules are kind of made to be broken and/or drive ESL learners insane.

Judging from the excellent data set collected by @John, I believe it is a combination of two phonetic rules, one based on manner of [l_] vs. [ɹ_], and one based on voicing of the fricatives when followed by /i/ (I won't go into other vowels, but there is a pattern).

First rule: left over from OE, unvoiced fricatives to voiced when followed by a (high) vowel.

  • breath vs. breathing
  • grief vs. grieving
  • worth vs. worthy

Second rule: R-vocalization and r-colored vowels influence voicing in /'wɚði/. The influence of /l/ on the interdental fricative, however, is not sufficient to ease into a comfortable voiced pronunciation.

The slight difference in /l/v and /ɹ_/v -- liquid vs. rhotic -- may be enough to trump the First rule based on the influence of the rhotic /ɹ_/ on the fricative. Therefore we see a voiceless "th" when preceded by /l/.

That there are so few -orthy words (c.f. all the words that can add "-worthy" to make an adjective), but with the limited data set and with the very relevant OED citation from @Alex with "worthy" making an appearance long before "-lthy," later to be followed by "wealthy..." @John, I think I see where you're going....

Both /l/ and /r/ are alveolar and located very close in the mouth, but with the slight difference in pronunciation when it comes to approximates. Interesting to note that in languages like Japanese, which does not differentiate these phonemes, many native speakers do not notice the difference between the pronunciation of /l/ and /ɹ/ in English.

Could this be the case of the Great Vowel Shift?

Since the meanings of the two words are so closely related, I checked the etymology of wealth:

From Middle English welth, welthe, weolthe ("happiness, prosperity"), alteration (due to similar words in -th: compare helth ("health"), derth ("dearth")) of wele ("wealth, well-being, weal"), from Old English wela ("wealth, prosperity"), from Proto-Germanic *welô (“well-being, prosperity”), from Proto-Indo-European *wel- (“good, best”), equivalent to weal +‎ -th. Cognate with Dutch weelde ("wealth"), Low German weelde ("wealth"), Old High German welida, welitha ("wealth") Related also to German Wohl ("welfare, well-being, weal"), Danish vel ("weal, welfare"), Swedish väl ("well-being, weal"). from here

and worth:

From worth or wurth, from Old English weorþ, from Proto-Germanic werþaz (“towards, opposite”) (the noun developing from the adjective). Cognate with German wert/Wert, Dutch waard ("adjective"), Swedish värd.

Keep in mind that "þ" (thorn) evolved into "th." One might just toss the potential rules at this point.

Always handy: IPA chart

Linking R

Your Answer

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