2021-11-13
• no tags • 930 words
My first ever blog post!
The /s/ and /r/ phonemes in Germanic have some quite interesting alternation between languages that you may not have even noticed.
For instance, in German, the word to be, sein, is conjugated in the past tense with an /r/: war, warst, waren, but in the past participle this /r/ somehow turns to an s (/z/): gewesen. And in Dutch, the same verb zijn cannot make up its mind: ik was becomes wij waren in the plural, and then the participle goes back to having an s: ik ben geweest. In fact, the same happens in English if you think about it: I was, but you were.
A similar phenomenon happens in English. We have lose and lost, but then we have the word forlorn. Similarly in Dutch, the word for lose is verliezen, which then somehow becomes verloor in the past tense. But then in German, we have verlieren and verloren where both have /r/. What is happening here?
The answer lies in two important sound changes that occurred in the history of Germanic - the most important here being Verner's Law. The other one, which you may have even heard of, is Grimm's Law (named after Jacob Grimm, who is the same person who wrote the fairy tales).
Grimm's Law showed a set of correspondences in sounds between German (and other Germanic languages) and other Indo-European languages, such as Latin and Greek for instance. For example, Latin p corresponds a lot of the time to a Germanic f: Latin pes and Greek pous, compared with English foot. Or Latin piscis compared with English fish. I'm sure if you think about it there will be many more you notice.
The full list of correspondences is quite long, but for some more examples:
So how does Verner's Law come into this? And how does this explain s and r? After all, none of these changes above feature s at all, let alone r.
The problem is that Grimm's Law does not explain everything. There are unfortunately exceptions to Grimm's Law, which causes a bit of a problem for linguistics, especially those who want to appear scientific: it is not really seen as good for a scientist to act like his theory explains everything, when in reality he is just picking and choosing the parts where his theory works and using that as "evidence" that he is correct.
And so to explain these exceptions, Karl Verner came along and proposed a new sound change that meant that the sounds did not just change into one sound, but two! So p for instance could change into one of /ɸ/ (f) or /β/ (v) depending on the situation.
The main conditioning factor for this change is stress. So in the case of two very common words, for instance Vater and Bruder where the t or d sound ideally be the same according to Grimm's Law (because the PIE forms both have a /t/: bʰréh₂tēr (meaning brother) and ph₂tḗr (meaning father)) so as in the sound changes I listed above, these should both have a /þ/. But it is clear that the fact that the German forms have different sounds - /t/ and /d/, means that they would have not both had a /þ/, but must have had different sounds there.
As can be seen in the above PIE forms though, the stress is different (in PIE, stress is marked with a acute accent). And so because of this difference in stress, the resulting forms end up with different sounds.
Where /s/ and /r/ come into it, are because Verner's Law also affected /s/. It is worth pointing out that /s/ was not actually affected by Grimm's Law, but it is by Verner's Law. This caused the /s/ sound to alternate with a /z/ sound. And the final piece of the puzzle is that /z/ underwent a process called rhoticisation, where a sound becomes rhotic, or r-like. And so /s/ and /r/ ended up weirdly alternating in a lot of Germanic languages, causing these weird cases where was becomes were, or where you are lost in English but verloren in German.
The final piece of the puzzle is a process called levelling. Languages often undergo a regularisation, where odd or irregular, or generally difficult to remember paradigms become more regular. In fact, I am sure you can think of some examples of levelling you have heard in your day-to-day life: people might say they swelledas the past tense of swell (as opposed to swoll) and many examples of this have occurred in say, the transition from older forms of English to modern-day English: we speak of cows not kine, and some are partway through transition: we may varyingly say octopusses or octopi, instead of octopodes.
So in this case, the irregular variation between s and r was levelled in a lot of cases to be a bit more regular. Taking lose as an example, in English we levelled it all to /s/ (lose, lost) whereas in German they levelled it all to /r/ (verlieren, verloren). In Dutch, they did not level it completely, and ended up with the present tense verliezen and the past tense verloren. A similar thing happened with all the other examples: in English, was and were did not undergo levelling, but in German they did, yielding war and waren. But the past participle in German did not get levelled, giving gewesen.
As a side note, the commoner a word is, the less likely it is to be levelled. This is because the levelled form is used more often and so is more likely to be corrected.