Written by Dyami Millarson
Today I woke up with an etymological rumination which I might as well write down because it does not matter how a thought process starts, but what matters is ultimately what it leads to: Latin Chērūscī and Ancient Greek Χηροῦσκοι (chēroûscoi) may be derived from Proto-Germanic hēr- hair and -isk- -ish in the same way that Hindeloopen Frisian méénsk, German Mensch and Dutch mens come from Proto-Germanic mann- man and -isk- -ish. So, if we make this connection, the long vowel of the Latin Chēr- and Greek Χηρ- (chēr-) is matched with a long vowel in Proto-Germanic; the long vowel of the Germanic match which is being considered particularly resembles the phonetic quality of the long vowel -η- (-ē-) in the Ancient Greek language of classical times.
Given that technological names, such as Latin Francī Franks and Saxōnēs Saxons, seem to be a Germanic tribal nomenclature tradition, is there any indication that naming a tribe after hair might be a Germanic tradition as well? Is there anything to suggest that naming a tribe after hair is considered suitable in Germanic culture? In fact, there may be a parallel: Hasdingī, which may likewise be derived from Proto-Germanic hazd- hair and the suffix -ing-. Therefore, Hasdingī seems a very similar formation.
Why may hair have been culturally important to the Germanic peoples? Just like modern people who wish to show they belong to a certain social group or ‘tribe’ with their hairstyle, the Germanic tribes also had different tribal hairstyles, which could be used to distinguish the tribes. For example, the Suebians had the Suebian knot (if you are interested in this topic, see this article).
Are there any problems with the Proto-Germanic reconstruction Hērisk- based on the Latin and Greek forms? Does the reconstruction hold up to diachronic phonological scrutiny?
The Latin and Greek forms are trisyllabic. Let us analyse the syllables. Lewis and Short include three different Greek forms in their Latin dictionary: “Chē̆rusci, ōrum, m., = Χηροῦσκοι. Strabo; Χερουσκοι, Dio.; Χαιρουσκοί, Ptolem.” The vowel difference in the first syllable of the three listed Greek forms is interesting. To know the exact pronunciation of each of these three first syllables, we need to periodise them, and in order to do so, we must identify the authors:
- Strabō, who wrote Χηροῦσκοι (Chēroûskoi), must be identical to the Greek Στράβων (Strábōn), who lived in the period between the 1st century BC to the 1st century AD.
- Dio. and Ptolem. are not included in the list of abbreviations pertaining to the Lewish and Short dictionary of the Latin language.
- Page 190 of Germanien und seine Bewohner identifies Dio. with Dio Cass., so the author, who wrote Χερουσκοι (Cherouskoí), must be Diō Cassius, whose Greek name was Δίων Κάσσιος (Díōn Cássios); he lived in the period between the first and second centuries AD.
- The intended Ptolem., who wrote Χαιρουσκοί (Chairouskoí), may be identified with Claudius Ptolemaeus, whose Greek name was Πτολεμαῖος (Ptolemaîos); he lived in the period of the 2nd century AD.
We may now conduct a diachronic phonological comparison of the three first syllables based on the period in which the three authors lived:
- Strabo’s Χηρ- (Chēr-) may, in fact, have been pronounced short; so, instead of the expected elder Attic realisation /kʰɛːr/, Strabo’s realisation may have been /kʰer/. By the looks of Χηρ- (Chēr-), one may reason that Strabo’s /kʰer/ goes back to /kʰɛːr/, but that presupposes that the Germanic tribal name was already adopted into Greek at an earlier date, while Strabo’s form may actually also have been a recent adoption, probably from Latin, which would in that case suggest that the Latin syllable should not be Chēr- (long) but Chĕr- (short).
- Dio’s Χερ- (Cher-) must have been realised as /kʰer/, which is identical to the realisation of earlier times, and which may be identical to that of Strabo, even though Strabo spelled it in such a way that one might suspect the syllable was pronounced /kʰɛːr/ in earlier times.
- Ptolemy’s Χαιρ- (Chair-) must have been pronounced with a short monophthong; so, instead of the expected elder Attic realisation /kʰair/, Ptolemy must have realised the syllable as /kʰɛr/, which would be (almost) identical to the realisation of Latin Chĕr-.
Lewis and Short were apparently in doubt about the length of the monophthong of the first syllable of the Latin form, as they wrote Chē̆r-, wich means the vowel could either be short or long. The evidence acquired from the periodisation of the Greek forms and their pronunciations seems to yield a short rather than a long monophthong in the first syllable of the trisyllabic Latin form. Consequently, we may reconstruct the Latin syllable as Chĕr- instead of Chēr-, the latter of which I had initially reconstructed on the basis of the classical pronunciation of Greek Χηρ- (Chēr-).
The -ου- of the second syllable of all three Greek forms was in all likelihood realised as a short monophthong /u/ unlike in earlier times where it was a long monophthong, therefore suggesting a short -ŭ- in the Latin form as well; I had initially reconstructed a long vowel on the basis of the elder pronunciation of Greek -ου-.
The -οι- (-oi-) of the last syllable corresponds to the Latin -ī-; they are both inflectional endings, and therefore they do not matter as much for our revonstruction. We should, however keep in mind that the aforementioned endings indicate that the word is in the plural, and therefore, if we reconstruct the full word with the inflectional ending, we may also put it in the plural to match it with the attested Greek and Latin forms.
Based on what the Greek forms appear to suggest, we may conclude that the Latin form should probably be Chĕrŭscī with the vowels of the first two syllables being short and only the inflectional vowel of the third syllable being long. Using Chĕrŭscī, we may initially reconstruct Proto-Germanic *heruskōz and Gothic *heruskōs. The Proto-Germanic and Gothic ending with -sk- is normally not -usk- but -isk-, however, and so one may be tempted to correct it to *heriskōz ~ hiriskōz for Proto-Germanic and *heriskōs for Gothic, both of which would mean “those who are of the sword.” If we are not satisfied with correcting the vowel of the second syllable in the Latin and Greek forms to -i- and wish to follow it more closely by preserving the -u- without breaking the phonological rules of Proto-Germanic and Gothic, we may come up with the Proto-Germanic *herutiskōz ~ hirutiskōz and Gothic *herutiskōz, which would have one more syllable than the Latin and Greek forms, and which would mean ‘those who are of the hart.’ Yet, the deletion of one syllable in Latin and Greek would seem understandable considering that *herutiskōz ~ *hirutiskōz is somewhat long and hard to pronounce; Romans might have found it complicated to say *Cherutiscī and they may have simplified it to Cheruscī. It is even thinkable that the Cheruscans themselves used *heruskōz ~ *hiruskōz as a simplified colloquial form of *herutiskōz ~ *hirutiskōz; given that heru-skôz ~ hiru-skōz ‘those who are of the hart’ could not be confused with her-iskōz ~ hir-izkōz ‘those who are of the sword,’ it was definitely possible to simplify *herutiskōz ~ *hirutiskōz to *heruskōz ~ *hirutiskōz. Similar simplifications occur even in Dutch today, and Germanic languages do have a tendency of losing unstressed syllables, so it would not be entirely surprising if the -ti- were left out in colloquial speech. Moreover, the deletion of syllables was not unheard of in Latin and Greek poetry, so why should it be surorising that Germanic people might also delete syllables, perhaps in colloquial and poetic contexts?
Why did I reconstruct -e- ~ -i- for the first syllable in Proto-Germanic? When reconstructing Proto-Germanic words, why may one religiously apply the e > i sound change in certain environments? Vladimir Orel would reconstruct -e-, while Don ringe would reconstruct -i-. The former is the underlying vowel, but the latter is the vowel which is yielded from the influence of /i/; the change from underlying -e- to -i- is called i-umlaut. On page 123 of the work From Proto-Indo-European to Proto-Germanic (2006), Don Ringe explains the evidence for the Proto-Germanic e > i as follows:
In fact i-umlaut occurred in [pre-Proto-Germanic] [...]; in Norse and northern [West Germanic], i-umlaut occurred in other vowels before the loss of most unstressed *i. Much of our evidence, then, will consist of [Old Norse] and [Old English] forms which have endings that contained *e in [Proto-Indo-European] but which also exhibit i-umlaut, thus demonstrating that those *e had been raised to *i in [Proto-Germanic].
On this basis, one may assume that the e > i sound law was still operational during the time when the indigenous name of the Cheruscans was formed. Given that the underlying e would correspond to a superficial i in the case of *her- ‘sword’ and *herut- ‘hart’ due to being compounded with the suffix *-isk- which contains the necessary /i/ to bring about such a vowel shift, I must say that the *Hir- of both *Hiru(ti)skōz and *Hiriskōz are less similar to Latin Cher- than *hēriskōz, which would not be subject to the previously described i-umlaut. The Latin short e /e/ is more open than the Proto-Germanic short e /e/, but Proto-Germanic long ē is more like the Latin short e /e/ in terms of openness, so although vowel duration does not match between Latin short e /e/ and Proto-Germanic ē, it does match in terms of vowrl quality. Likewise, the seeming discrepancy between Latin -u- in the second syllable and Proto-Germanic -i- may not be such a problem when we consider that the Latin short i was more open than the Proto-Germanic short i; the Latin short i was, in fact, like the Modern English i as in bitter, while the Proto-Germanic i was like the Modern Italian i in città — a reverse situation which might seem unbelievable at first. Since the Germanic i was not a match with the Latin i anyway, the Romans may substited it with the Latin u for a number of reasons:
- They may have failed to recognise the Germanic short i as an i and simply looked for a sound they thought was equivalent. Also consider the fact that the i and u were — in very specific cases — interchangeable in the written Latin of Roman times due to the fact u/i could be used to render a sound intermediate between i/u. Take, for example, optimus/optumus, where that intermediate vowel occurs in the second syllable.
- The Romans may have misheard the Germanic i as an u; they may genuinely have thought they heard an u. A possible explanation for this is that syllable stress may have played a role in how accurately the Romans heard the sounds: the first syllable of the Germanic word would have been easier to hear while it was strongly stressed (i.e. primary stress), and so the vowel of the Latin Cher- should be more reliable than the vowel -u- of the second syllable, which may be based on a weakly stressed (i.e. secondary stress), though not unstressed, -i- in Proto-Germanic. Another possibilities for this is that information changes as it passes through many different persons, as may be seen modern word games where one person in a circle says one thing to their neighbour in closewise direction and the word is passed on until the last person — the more people involved, the more different the final result will be. I often hear my fathet saying in idiomatic Dutch: “Als iets langs een heleboel schijven gaat, dan gaat het vaak mis.” (If something passes by layers, i.e. persons, it often goes wrong.) Therefore, if the Romans got their information indirectly, all sorts of phonetic discrepancies may have creeped in. For instance, we may suppose that intermediaries, who told the Romans the name of the indigenous tribe, may have changed the Germanic vowel i to a sound that the Romans heard as u.
- They may simply have inserted the u because it sounded good to their ears. So it could all have been for reasons for euphony.
The Romans, who transferred the name of the Cheruscans into Latin, were not modern phoneticians after all; they were just trying to roughly match sounds, it was a mere approximation. So, we should perhaps not be taking every slight difference too seriously; the differences do matter somewhat, but they should not really stand in the way too much for supposing that -uscī might actually be equivalent to *-iskōz instead of *-uskōz in Proto-Germanic. Therefore, purely based on the Latin evidence, *Hēriskōz does seem plausible, and considering the vowel of the Latin Cher-, it does seem more plausible that Cheruscī corresponds to Germanic *Hēriskōz than *Hirutiskōz or *Hiriskōz; in other words, it does seem that *Hēr- is a more likely candidate than *Hir- when judged by the vowel quality of the Latin short e, noteithstanding that the Latin short e is even more open than usual before the r as in the case of Cher- if we analyse the stem vowel as short. However, if we analyse the stem vowel as long, namely Chēr-, then the best match would again seem *Hēr- instead of *Hir-. If the first syllable of the Latin form were Chī̆r-, so either Chĭr- or Chīr-, then I would definitely connect it with *Hir- since Latin -i- is more closed than -e-, making it a prime candidate for rendering the stressed — therefore clearly audible — Proto-Germanic -i-. Why might the Romans have rendered Proto-Germanic *Hir- as Chir- but -isk- as -usc-? We must bear in mind that there is a difference between the 1st syllable and second syllables in terms of stress in Germanic. While the Romans would the second syllable of our hypothetical form of the ethnonym, namely Chirúscī, the speakers of Germanic would have given primary stress to *Hir- and secondary stress to -isk-, thus making *Hir- sound relatively stronger while -isk- relatively weaker. As a result of this state of affairs, the Romans may have heard/analysed/interpreted the two sounds differently.
Since Gothic has herdīs for the expected hirdīs and there are many more such examples in Gothic where short e occurs before r instead of the expected i which is the result of i-umlaut, is it possible that the Germanic i sounded differently before r? Is it feasible that Germanic had something akin to the Latin or Swedish /æ/ before /r/? Given that the sound apparently became more open in those positions in Gothic, one may be attracted to that notion; nevertheless, other Old Germanic languages point to the retention of umlauted i before r, and so it rather seems to be a development peculiar to Gothic. In conclusion, Gothic does not provide strong support for the notion that *Hirutiskōz or *Hiriskōz should be *Herutiskōz or *Heriskōz in Proto-Germanic. It does, however, prove that such a sound development could take place. Should we then consider that the Cheruscans had a similar sound change as the Goths? Probably not, because Gothic is from a later period; the Latin and Greek forms are attested during the pre-Gothic, namely Proto-Germanic period. Regional differences in pronunciation are always possible, but it is the safest to assume that if the Cheruscans were speakers of Germanic, then they must have spoken the Proto-Germanic that we can reconstruct based on descendant languages. After all, we may assume that the Proto-Germanic that the Cheruscans spoke must by and large have corresponded to the language we can reconstruct, and we may therefore also assume they possessed the Germanic vocalic shift from e to i as a result of the i-umlaut. We may make such assumptions because of the time period; the further we go back, the more alike the local forms of Germanic must have been. Since the attested Latin and Greek forms of the ethnonym of the Cheruscans stems from an earlier time period than the bulk of our materials of the Gothic language, we may assume it was way more like the general Proto-Germanic in terms of phonological characteristics.