This post is a brief introduction to the process of debuccalization, a sound change with the power to dispel confusion in various languages of Europe and beyond. With the help of some concrete examples taken from three languages, this is a concept that I believe might come in rather handy for language learners.
As is often the case, its name, which may at first glance seem strange and over-technical, nicely summarises the process. It comes from Latin de– ‘from, un-‘ and bucca ‘mouth’ (the origin of French bouche). To put this into more familiar English terms, debuccalization is when a a particular sound stops being produced in the mouth and over time has its place of production moved to somewhere else – specifically the glottis in your larynx.
This is the place where sounds like [h] (as in English hat) and [ʔ] (as in English uh-oh) are made. Debuccalization has occurred in countless languages, modern and old, in and outside the Indo-European family. This article takes examples from English, Old French, Welsh and Ancient Greek.
In English, the glottal stop [ʔ] is a common sound across many varieties of the language and some of its occurrences have come about through debuccalization. For example, a well-known feature of Essex English is that where Standard (read: prestigious) British English has [t] between vowels, Essex English has [ʔ] instead – imagine wa’er and bu’er for water and butter. This change from [t] to the glottal sound [ʔ] is still ongoing and the feature is common today across the South-East of England and parts of Scotland.
The example of Essex English shows us that debuccalization can be limited to certain positions in the word or syllable – you will not find [ʔ] in place of [t] at the start of words in Essex English. Importantly, where debuccalization occurs in one language may not be where we find it in another.
In Old French, the consonant [s] often underwent debuccalization. Originally produced in the mouth (with the tongue behind the teeth), [s] became [h], but only when positioned at the end of a syllable (the coda) and before another consonant. The resulting [h] in time disappeared, though traces of it may remain in the preceding vowel and in the spelling of the word.
For example, French signer ‘to sign’ and visiter ‘to visit’ have kept their [s] (inherited from Latin sīgnāre and vīsitāre), while arrêter ‘to stop, arrest’ has lost it, because the [s] in its Latin ancestor restāre comes before the consonant [t]. Examples of this abound and we can set them beside English sister words that retain the [s].
|Latin Origin||English Descendant||French Descendant|
Welsh is another language that shows the effects of debuccalization. The change something that sets Welsh apart from Irish, another Celtic language. In many words, Irish maintains the original [s] sound of the languages’ common ancestor. In Welsh, it instead became [h] at a prehistoric point.
The Welsh word for ‘old’ is hen (as in Yr Hen Ogledd ‘The Old North’, the historical Celtic-speaking region of northern England and southern Scotland). In Irish meanwhile, the word is sean. It is through the Irish word that we can better appreciate their connection to other ‘old’-words in other languages, such as Latin senex ‘old’, senior ‘older’ and also senatus, originally a meeting of elders.
Moreover, this sound change connects Welsh haul ‘sun’ to Irish súil ‘eye’, two members of a family of sun-words that also includes Latin sōl, Ancient Greek hḗlios, Russian solnce and German Sonne. Likewise, the same languages reveal a salt-family of words. There’s Latin sāl, English salt, Ancient Greek háls (the origin of halogen), as well as notably Irish salann and Welsh halen. The change also unites their third-person feminine singular pronouns (that is, meaning ‘she’), since Welsh has hi and Irish sí.
Debuccalization comes in handy for learning Ancient Greek. At the prehistoric stage of Proto-Greek, the sound *s became [h] at the beginning of words and in between vowels (and other resonant sounds). In the latter, the resulting intervocalic [h] later disappeared entirely, as in Old French, but word-initial [h] survives until the classical era. This debuccalization firstly allows us to find useful cognates of Greek words in Latin, in which this change did not occur.
Secondly, this also can explain variation in the grammar of Ancient Greek that might otherwise remain a bit of a puzzle. Take dôron ‘gift’, a noun of the neuter gender and the second declension. Since the plural of dôron is dôra, we can assume that the affix that is added to convey the plural number is -a. But what about neuter nouns of the third declension, like ethnos ‘group, nation’, for which the plural is ethnē in Attic Greek? Do we have to learn a second plural affix for verbs like these?
No! It is in fact still –a that makes ethnē plural, and so -a is the only neuter plural ending to be learned. The stem of the word is ethnes– and it is to this that grammatical endings are attached. When the plural affix -a was added, the [s] of ethnes– found itself sandwiched between two vowels and so became [h] through debuccalization. *Ethnesa became *ethneha.
When intervocalic [h] then disappeared, the two vowels could merge in Attic Greek, as per the usual rules of contraction – in this case *ethnea became ethnē, marked as plural by the long vowel -ē. We can see the same processes across the various different forms of ethnos.
|Case and Number||Stage 1||Stage 2||Stage 3||Attic Form|
The same pattern works too for similar, common third-declension words, such as génos ‘offspring’ and étos ‘year’.
Understanding these changes to *s reveals wonderful connections between Greek and Old English, a language that is distantly related to Greek and that retains the inherited [s] sound. Not only do both languages make use of a definite article (that is, the), the words used for it come from the same prehistoric origin. Compare the singular forms for the definite article in Old English (on the left) and Ancient Greek (on the right), organised according to gender (masculine, feminine, neuter) and case.
What then are the similarities? One is that Greek t corresponds precisely to Old English þ, which usually stands for the sound [θ]. Here Greek preserves the original sound, while Old English shows the predictable outcome of Grimm’s law. We can also see that the neuter nominative forms in each language (þæt and to) are identical the neuter accusative (þæt and to), and that the masculine accusative forms in both languages contain the nasal consonant [n].
But what about debuccalization? Its effects are there in Greek ho and hē; the original *s, still present in their Old English counterparts se and sēo, has become [h] at the beginning of these two articles. Moreover, we can assume that the [s] in the Old English masculine and neuter genitive forms þæs and þæs would have at one point corresponded to a [h] in the Greek words as well, but, being trapped between two vowels, that [h] has since disappeared, leaving us with tou and tou.
Three examples of debuccalization from four languages – English, Old French, Welsh and Ancient Greek. As British English shows, the change can be vibrant and as yet unsettled, while, as Greek shows, it can also be a foregone conclusion, a mischief-maker long since lost in a muddle of grammar. What debuccalization is for certain, however, is regular, and if you are prepared to find the rule, it can only be of great help.