Secrets of the Words You Know
Blog 30 Labio-Velars 1 *gʷ and the secrets of a diabolical Parliament
The Labio-Velars – so often the source of surprises!
These sounds are by their nature complex, as they combine two articulations, one of the lips and one of the soft palate.
As a reminder, here is what we said in Blog post 11 about how they may originally have been articulated:
- “Of course we cannot know exactly how sounds were produced 5,000 or so years ago, but we have reasons to believe that the PIE labio-velars would have been rather different from, say, the initial ‘kw’ sound of ‘quell’, which is a close parallel to the consonant clusters ‘tw-’ or ‘dw-’ in ‘twelve’ and ‘dwell’.
- The labio-velars were perhaps closer to the West African sound often transliterated as ‘gb’, for example in the word ‘Igbo’, referring to the people and language of that name rooted in South Eastern Nigeria. To make this sound, your lips and the back of your mouth close and release simultaneously, whereas with ‘quell’ the velar ‘stop’ is released just before the breath reaches the lips and your lips articulate only a ‘w’ sound, and not a ‘b’.”
Their evolution is also less straightforward than many other stop consonants. For a start, in the ‘satəm’ languages like Indo-Iranian and Balto-Slavic, they lose the labial element, and merge with the plain velar stop consonants, which we have been discussing in the last three Blog posts. In the ‘centum’ languages, as we shall see, the evolution of these sounds is also in the direction of merger, but with more diverse results. In some cases the labial element lapses, but in others it is the velar that can vanish.
In Greek, though labio-velars were still written with unique syllabic signs in Linear B, they had completely vanished by merging into other pre-existing consonants before our earliest alphabetical records of Classical Greek, a few hundred years later. Greek also has a particularly surprising set of outcomes, which includes in some circumstances a change to a dental, which is neither a labial or a velar!
In English-inherited words, by contrast, some of them are still visible in writing to this day in a few inherited words beginning with ‘qu’ and most (but not all) of those beginning with ‘wh’. Some dialects, though not standard English, still indeed pronounce ‘wh’ with the aspiration shown in writing.
Labio-velars are often involved when you see some of the less obvious equivalences between words of common Indo-European heritage.
One development is found across most Indo-European languages groups that conserved labio-velars and across all three PIE labio-velars*gʷ, *gʷh and kʷ, so we will mention it here:
- before or after a ‘u’ or ‘w’ the labial element disappears, leaving just the velar element.
This means that the outcomes will be simple velars, which will develop exactly like the equivalent sound ‘*g’, ‘*gh’ or ‘*k’ as the case may be. (Such an outcome parallels what happens to all labio-velars in any phonetic environment in the ‘satəm’ languages.)
This development is another example of ‘dissimilation’ (Box 7, Technical Note 3), and it is quite a natural one between a lip-rounded consonant and a lip-rounded vowel. You can imagine that ‘gʷu’ (and also ‘gʷhu’ and ‘kʷu’) are not stable sound-combinations: no words starting with ‘qu-u’ or ‘qu-oo’ exist in standard English, and if you repeat them quickly, you can feel how readily the ‘w’ element can be swallowed up.
Such dissimilation is sufficiently widespread to have gained the status of a ‘rule’: the so-called ‘Boukolos Rule’, from the Greek word for ‘cowherd’, where the ‘k’ is indeed a ‘de-labialized’ PIE *kʷ immediately following a *u. We will discuss the first element of the word ‘Boukolos’ shortly, and refer to the second element under *kʷ in Blog post 32.
Voiced: *gʷ
(i) *gʷ at the start of a word
In the case of *gʷ, we would expect, in line with Grimm’s Law, that Proto Germanic would show an unvoiced *kʷ in initial position, which would then yield a ‘kw’ sound in English. And indeed, this is what we typically find, with the spelling ‘qu’[1].
Here are some examples of initial *gʷ:
- The PIE root *gweiH- meaning ‘to live’, appears in English as ‘quick’, conserving its original meaning in the fossilised phrase ‘the quick and the dead’, but having of course evolved to take the meaning ‘fast’ or ‘rapid’. The Latin reflexes of this root, which we can see in borrowings like ‘vital’ and ‘vivid’, show that Latin did not conserve *gʷ in initial position before a vowel (unlike *kʷ, as we shall see shortly). Instead it lost the velar element and merged with the descendant of PIE *w. It then developed (as with inherited *w) to a ‘v’, as explained in Blog post 16 on *w.
In Greek, not only does *gʷ evolve to a plain velar before a *u, as explained above, but to a labial before ‘a’, ‘i’ and ‘o’ and – remarkably – a dental before ‘e’ but also ‘y’. We can show both of these changes in English borrowings from Greek deriving from this particular root, which also occurs with metathesis (Technical Note 3, Box 1) as *gwyeH- (giving *gwyō- as the laryngeal was, as you will now see, o-coloured):
- dentals from the metathesized root version *gwyō- in ‘zoo’, ‘zoology’ etc (Greek ‘z’ probably representing a ‘dz’ or ‘zd’ sound, as we saw under *d). Ancient Greek vowels are mostly faithful representations of the assumed vowels of PIE, taking account of the disappearance of laryngeals;
- labials from the zero-grade of *gweiH, that is to say*gʷiH-, which we see in the large number of borrowed or made-up words with the element ‘bio’, from ‘biology’ to ‘amphibian’ and ‘anaerobic’.
The fact that the first elements of ‘zoology’ and ‘biology’ come from the same root shows rather starkly how different grades of root with a labio-velar can produce seemingly quite unrelated items, even within a single language. Taking Ancient Greek as an example, it seems that to Agamemnon and his contemporaries ‘zoology’ and ‘biology’ would have started with the same sound (as shown by the Linear B script), whereas to Socrates and his contemporaries, a few hundred years later, the two words would have seemed entirely unrelated, just as they do to us.
- The PIE stem *gʷreH-u means ‘heavy’. It has supplied English with a good set of borrowings, such as ‘grave’ and ‘gravity’ from Latin (showing that before an *r, *gʷ evolves in Latin to a simple ‘g’), ‘barometer’ and similar scientific terms drawn from Greek, and ‘guru’ from Sanskrit, the ‘weighty’ person being an epithet for a teacher. (As expected, Sanskrit, which like all ‘satəm’ languages merges the labio-velars with the plain velars, maintains the *g before the ‘back’ vowel ‘u’.) So where is the inherited English equivalent? Well, it’s that rather rare word ‘quern’, the traditional grinding stone for grain, and named from its weight as ‘the heavy one’.
- A very similar-looking root *gʷerH- has the core meaning of ‘praise’. There is no direct English descendant, but it is the source of the Celtic-derived word ‘Bard’ showing that Celtic has evolved *gʷ>b. The final ‘d’ of this word is from *dheH- (‘to put’), which we saw under *dh, so that a Bard is someone who ‘puts forward’ praises.
The root also lies behind English borrowings from Latin like ‘grace’ and ‘gratitude’, again showing *gʷ giving Latin ‘g’ before ‘r’.
- The PIE root *gʷou- gives us, in accordance with its original meaning, the word ‘cow’, where the velar element has lost its voiced character, in accordance with Grimm’s Law, and the labial element has disappeared because in Old English the following vowel was a ‘u’.
You would expect from the words for ‘life’ above that the Latin for ‘cow’ would also begin with a ‘w’-sound before the vowel, and that this would give us borrowings beginning with ‘v’. However – and this may well be a case of a dialect word creeping into the Latin mainstream – our ‘cow-related’ words inherited directly from Latin like ‘bovine’, or via French like ‘beef’, all show that the Latin for ‘cow’ starts with an unexpected ‘b’.
Greek, on the other hand, would be expected to show a ‘b’, and does so, as the name of Alexander’s horse ‘Bucephalus’ (literally ‘ox-head’) shows. ‘Butter’ (literally ‘cow-cheese’) is also from Greek. A Celtic example, similarly confirming *gʷ>’b’, is the fine mountain by Glencoe in Scotland known as ‘Buachaille Etive Mor’, or ‘The great [mor] cow-herd [Buachaille] of Etive’.
Sanskrit and its descendants show the expected ‘g’, as in ‘Gurkha’, with the literal meaning of ‘cow-protector’, as we saw in Blog 14.
- The English descendant of the PIE root *gʷem- has similarly kept the root’s meaning of ‘to come’, while losing the labial element. (The PGmc derivative would have been *kʷemana. In zero grade the *e would be lost and hence the labial element of the initial *kʷ was vocalised to ‘u’ in Old English ‘cuman’[2], thus removing the labial element of the initial consonant.)
In this case, Latin does show the expected initial ‘v’, as in our many borrowings from ‘advent’ (the ecclesiastical season of ‘coming’), ‘circumvent’ (come round something) or even ‘convenient’ attest. Greek again offers us words and word elements beginning with ‘b’, such as ‘basis’, ‘diabetes’ (from the ‘running through’ of heavy urination) or ‘acrobat’ (‘walks on tiptoe or fingertips’). From the sub-continent, the first element of the word ‘Juggernaut’, a large vehicle carrying an image of Krishna, ‘Lord of the World’, also comes from a reduplicated form of this root (Sanskrit ‘jagat’: ‘all that moves’ so, ‘the world’). The first ‘a’ stems from a PIE *e, hence the fronting of the initial*g>j (pronounced ‘zh’).
It’s worth noting that the development of *gʷ> b in Greek and Celtic is a further way in which the ‘gap’ in the consonant system visible from the absence or near-absence of *b in PIE has been closed: there is indeed a tendency to maintain a degree of symmetry at least among stop consonants.
The sound code of English initial ‘qu’ can be set out as follows:
English initial qu = PIE*gʷ = Latin v (g before r) = Greek b (before a/i/o)/ t (before e)/z (before y) = Celtic b. PIE*gʷ = Balto-Slavic/Indo-Iranian g before original back vowels and palatalized sounds before original front vowels.
The same applies to the few cases like ‘cow’ and ‘come’ where a written ‘c’ goes back to a PIE*gʷ.
(ii) Secrets of a diabolical Parliament
For our example of wide-ranging meanings, we will take the root *gʷelH-, whose core meaning seems to have been ‘to throw’.
Certainly this is the case for Greek descendants of the root visible in borrowings like ‘ballistics’ or ‘parabola’, with the expected development of *gʷ>’b’ before all vowels other than ‘e’ and ‘u’.
Greek, like many other Indo-European languages, has a wide range of prepositions/preverbs that can radically adjust the meaning of the roots to which they are attached. This root provides an excellent demonstration. For example, with different prepositions the sense develops to include something thrown in (so ‘embolism’), changed (so ‘metabolism’) or thrown at someone in slander (as in ‘diabolical’, which also comes into English via late Latin as ‘devil’). ‘Hyperbole’, ‘problem’ and ‘symbol’ are other members of this cluster.
Similarly, a ‘parable’ is a way of explaining something by a story about something ostensibly different. This word, taken over from Greek into late Latin, became in some Romance languages the main way of describing speech, as we can see from borrowings like ‘parley’, ‘parole’ and ‘Parliament’ from Old French, where only the ‘l’ of the root element of ‘parable’ is left; or ‘palaver’ from Portuguese, where the ‘v’ is the legacy of the Greek and late Latin ‘b’, and we can see a neat case of metathesis of the ‘r’ and the ‘l’.
Finally, late Latin also takes from Greek a further sense of ‘to dance’, giving rise to English borrowings ‘ballet’, ‘ballad’ and ‘ball’ (as in a dancing party).
Is there an English inherited word from this root? ‘Quell’ is usually quoted as an example. It is true that its sense of ‘to kill’ in Old English, which has been softened over the years to ‘put down’, is quite distant from ‘throw’, but it appears to be a good match in terms of sounds.‘Quail’ in the sense of ‘shrink from’ is also quoted as a possible member of this family (in this case perhaps showing the effects of violent action). The word ‘kill’ itself is of uncertain origin, though some authorities do link it to this root as well.
(iii) What’s an inherited English word from another PIE root beginning with *gʷ?
The PIE root *gʷ enH-, with an original meaning of ‘woman’, is familiar from borrowings from Greek such as ‘gynaecology’ or ‘misogyny’, where the labial element is, as expected, lost before what would have been a ‘u’-sound in Ancient Greek. In Celtic, the expected ‘b’ appears in the first element of Irish ‘banshee’ (literally ‘fairy woman’).
What’s a noun that English has inherited from this root?
[1] If you wish to know the reason for this spelling, as opposed to old English ‘cw’, (and a whole series of alphabetical wonders) I recommend Danny Bate’s recent book, ‘Why Q needs U’ (Blink Publishing, 2005)
[2] Indeed, the pronunciation remains ‘u’: it seems that the reason for the written ‘o’ is to do with scribal conventions to make it easier to read than having a ‘um’ sequence in