Linguistic commentary from a guy who takes things too literally


Posted by Neal on February 10, 2005

Adam has an unusual ability for a four-year-old: He can say both ‘th’ sounds, [θ] (as in thick) and [ð] (as in this). It’s not so much that he’s precocious in this regard as that he tends to keep his tongue a little forward in his mouth in general. So even though he makes perfect [θ] and [ð], with his tongue right between his front teeth where it needs to be, he also keeps it there for his alveolar consonants (t, d, n, s, z), so that his perfect th’s come along with a lisp and funny-sounding t’s, d’s, and n’s.

Meanwhile, Adam has been playing Mousetrap a lot during the past week, and I get to hear his perfect [θ] whenever he rolls a three. But when he rolls higher than a three, say a five, he’ll pick up his mouse, and count:

One, two, three, Thor, thive!

Where other kids are busy turning three and free into homophones just because they can’t say [θ], Adam’s taking that [θ] and going to town with it! But why is he replacing his [f] with [θ], when he doesn’t have a problem saying [f]?

It’s because his three has contaminated them. Contaminated them, I tell you! And not only is it a clear case of contamination (i.e., semantically related words becoming more similar to each other phonetically), it’s even a canonical one. According to Hock’s Principles of Historical Linguistics, “contamination occurs frequently in antonyms….” and get this, “It is also commonly encountered in the numerals” (p. 197). The first example Hock gives is from Latin, where septem ‘seven’ and decem ‘ten’ influenced an earlier Latin noven ‘nine’ to become classical Latin novem. The second example is a dialect of Greek in which the initial [h] of hepta ‘seven’ contaminates okto ‘eight’, turning it into hokto. So Adam’s following the Romans and the Greeks (not to mention the Norse) when he says “Thor” and “thive.”

Just as linguists say you shouldn’t, I tried to correct him, to undo the contamination:

Me: Adam, say “three, four, five.”
Adam: Four, five!
Me: Say “three, four, five.”
Adam: Free, four, five!

D’oh! Now the four and five are ganging up on the three and contaminating it!

8 Responses to “Contamination!”

  1. ACW said

    Another couple of lovely examples: four and five both start with [f] in English due to exactly this sort of contamination. Indoeuropean had *kwetwer and *penkwe, whose distinct initial sounds ought to show up in English as wh- and f-. Clearly, the f- of 5 contaminated the wh- of 4 at some stage. (This trait is shared by all the Germanic languages, so the confusion must predate their dispersal.)

    Another example is in Slavic languages, where the word for 9 unexpectedly begins with a d- instead of n-. The d- is almost certainly contamination from the 10 word.

  2. Neal said

    Thanks, ACW, for the great examples!

  3. such said

    Hmm, interesting post.
    Forgive my ignorance, but I was wondering why the ability to say the “th” sounds is unusual in a four-year old.
    In India, the word “thatha”, meaning grandfather, with the “th” as in “that”, is one of the first words children say when they learn to speak. The softer “th” sound they say by about 2 or 3 years.
    Of course, some do have problems with the “f” sound, sometimes replacing it with “sh”.

  4. Neal said

    I don’t know *why* it’s unusual, but it seems to be the case, at least in English. I personally didn’t learn it until I was 5, and it is the last remaining pair of sounds for Doug (age 6) to acquire. I didn’t know it was otherwise in other languages until I read your comment. I wonder if someone’s written about it.

  5. […] it was again: appindex. Looks like Adam’s not the only one in our family to have produced a contaminated linguistic form. Two words with a phonetic […]

  6. […] readers may remember me talking about contamination with Adam’s counting, or with Doug’s pronunciation of appendix. And while I’m on the subject, here’s […]

  7. dw said


    Indians speaking English usually use dental stops rather than dental fricatives for the “th” sounds of English, because the fricatives are absent from almost all indigenous Indian languages.

    My guess is that children are producing the dental stops — perhaps you use them too.

    My two-and-a-half year old daughter can produce all the other consonant phonemes of English, but she still can’t pronounce either of the dental fricatives, despite being surreounded by adults who pronounce them.

  8. dw said


    Oh: I’ve just realized that you are referring to दादा – Hindi for “paternal grandfather”! That’s usually transliterated as “dada” or the like, which is why I didn’t recognize it written with “th”s.

    The Hindi pronounciation of that word uses the voiced dental stop, IPA [d̪], not the voiced dental fricative which is IPA [ð]. The latter sound is used by most native English speakers outside India — for example in the standard accents of the US, Canada, England and Australia. The dental stop is also found in most southern Irish speakers.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: