--- In
cybalist@yahoogroups.com, Piotr Gasiorowski <gpiotr@...> wrote:
>
> W dniu 2011-06-25 20:52, Joao S. Lopes pisze:
> > long or short O ?
>
> It's short in Latin.
>
> > Are there many examples of this initial cluster stl- in another IE
> > languages?
>
> No, that's why I argue for vowel syncope in the initial syllable. There
> are a few examples of stl- in Latin, though. Apart from <stlocus> we
> have OLat. stli:s > Class.Lat. li:s, -tis 'lawsuit', <stlatta ~ stla:ta>
> 'cargo-boat, barge' and the derived adjective <stlatta:rius ~
> stla:ta:rius> 'imported by boat' (both of which are rare poetic words
> preserved in their archaic form till Classical times), plus the clearly
> onomatopoeic hapax <stloppus> 'slap on an inflated cheek'.
I prefer to argue for /t/-increscence in Proto-Latin, like that which produced Germanic *str- from *sr-. And I think the morphology of <stlocus> can be explained like <focus>, as a back-formation from an /o/-grade implemental noun of a causative verb, this noun being reinterpreted in Latin as a diminutive. In my earlier post on <focus> I failed to address the gradation, so I will attempt to correct this oversight first.
Most Indo-European implemental nouns with suffixal *-tl- are zero-grade oxytones in *-tló-. These are typically neuter, but some masculines also occur. Thus we have *putlós 'implement of procreation, son' (Sanskrit <putrá->, Oscan acc. sg. <puklum>) from the root *peu- 'to propagate one's kind, procreate' seen also in Italic *powemom 'fruit' (Lat. <po:mum> 'id.', Umbrian <Puemune> dat. sg. 'to Pomonus').
One such implement not showing zero-grade is Lat. <po:culum>, Skt. <pa:tra-> 'drinking vessel'. Etruscan <putlumuza>, with the hypocoristic suffix -uza, clearly represents borrowing from *po:tlom in one of the older Italic languages. Since the Sanskrit accent is not given, the root-grade could be either *peh3- or *poh3-. But it seems more likely that this noun was also an oxytone, so /o/-grade can be inferred for the root.
In the IE languages which form a transitive verb 'drink' from *peh3-, the normal /e/-grade forms no tenses. Indic, Celtic, and Italic have a reduplicated present stem *pi-ph3e/o- (Skt. <pÃbati> 'he drinks', Old Irish <ibid> 'id.', Lat. <bib-> by assimilation from *pib- after the analogy of other reduplicated verbs). Greek uses a stem with /i/-infixation of the zero-grade, *pih3- (impv. <pîthi>, aor. <épion>, fut. <pÃomai>, <pioûmai>, nasal present <pÃ:no:>). Sanskrit has the same stem in the passive participle <pi:tá-> 'drunk'.
Ernout and Meillet noted that Lat. <caedere> 'to make fall, slay, cut down, cut' is effectively the causative of <cadere> 'to fall'. (The supine stem ca:su- < *ca:ssu- < *ka:d-tu- < *k^éh2d-tu- has normal grade; the Sanskrit perfect <s'as'a:da> has /o/-grade, and the present stem <s'ad-> is evidently analogical, since zero-grade would yield *s'id-.) Most IE lexicographers have rejected this connection, but <caedere> is not easily explained otherwise, and 'cut' need not be this verb's primary sense. A fragment of the Twelve Tables reads "si nox furtum faxsit, si im occisit, jure caesus esto" where the legal issue is not cutting, but slaying. Other clear-cut examples of causatives formed by /i/-infixation of zero-grades are hard to find, but if <caedere> from *k^h2id- is accepted, it is reasonable that *pih3- is parallel. The simple sense of *peh3-, whose causative means 'drink', is not immediately clear. One possibility is suggested by English <down> 'drink rapidly': "He downed a glass of milk." This is effectively the causative of 'go down' or 'be down'. Whatever *peh3- meant as a simple verb, it was most likely an intransitive of motion or station. Since it did not mean 'drink', a zero-grade implement *ph3tlóm would not have meant 'drinking vessel'. We can tentatively extract a rule for this situation: An implemental noun in *-tló- belonging to the causative of an intransitive root takes /o/-grade, not zero-grade of the root, whether or not an /o/-grade causative is directly attested.
The following two paragraphs are from my earlier post on <focus>:
[Rather than taking Latin <foculus> 'brazier, small stove, small altar' as the historical diminutive of <focus> 'fireplace, hearth', I regard <focus> as back-formed from <foculus>, which was reinterpreted as a diminutive although formed originally with the implemental suffix *-tlo-. The IE root is *dHegWH- 'to be warm' seen in the Latin causative <foveo:> 'I make warm'. The /o/-grade noun *dHogWH-tlo- 'implement of burning or warming' underwent assimilation to *dHoktlo-, becoming *foctlos in either Old Latin or Old Sabine. Cluster simplification then eliminated the /t/, without compensative lengthening because it was a mere stop. The resulting *foclos (whether inherited or borrowed) then became <foculus> through regular Latin phonetic development.]
[Evidence for parallel assimilation and cluster simplification comes from reflexes of *dHeig^H- 'to knead clay or dough, fashion, build'. Latin <fingo:> has the participle <fictum>, from the assimilated form *dHiktom. The ritual term <fi:tilla> 'sacrificial gruel' is evidently a Sabinism, a double diminutive of the P-Italic participial stem *fiht-. But Umbrian has <ficla(m)>, <fikla> acc. sg., <fiklas> gen. sg. 'sacrificial cake', which many scholars have referred to a nom. sg. *fig-kla: from earlier *fig-tla: with the feminine implemental suffix. However, a cake is the result of kneading and baking, not the implement thereof. Therefore I take <ficla(m)> as a diminutive of the participial stem with the same loss of /t/ postulated for *foclos above. That is, the Proto-P-Italic *fikta: 'kneaded mass, loaf of bread' etc. had the diminutive *fiktla: 'small loaf, cake' with its own diminutive *fiktenla: 'crumb, crumby broth, gruel'. Reduction of -ktl- to -kl- occurred before the P-Italic lenition of -kt- to -ht-. Thus *fiktla: provided our Umbrian nom. sg. *fikla:, while *fiktenla: yielded Sabine *fihtella:, Latinized as <fi:tilla>.]
Similarly, I think we should take <loculus> 'coffin, sepulchral niche' and <loculi:> pl. tant. 'coffer with compartments, schoolboy's satchel' not as diminutives of <locus>, but as continuations of an implemental noun from which <stlocus> was back-formed. The sense of *stloculus could have been, broadly speaking, 'compartment', which is an implement of apportioning things into their proper places. A coffer or satchel with many compartments resembles (somewhat fancifully) a sepulchral vault containing many niches for cinerary urns, and such a vault is so similar to a dovecote (containing many pigeon-holes) that it was termed <columba:rium>. From the viewpoint of the living, a sepulchral niche is also a rather small place, so if *stloculus was used (perhaps proverbially) in the sense 'small place', extraction of <stlocus> 'place' would be straightforward. I have no decent explanation for the distinction between the regular plural <loci:> 'separate places, passages in books' and the neuter <loca> 'connected places', arguably a plurale tantum 'neighborhood, region'.
The difficulty now is in finding a suitable verbal root attested or implied elsewhere in Indo-European. Such a root must begin with *sl-, but due to assimilation with the implemental suffix, it could end in *-k(^), *-k(^)W, *-g(^), *-g(^)W, *-g(^)H, or *-g(^)WH-. The normal grade should be intransitive and mean something like 'be in its proper place, fall into place, be properly assigned'.
Nothing semantically obvious jumps off the page. For the moment, my best guess is a root *sleg(^)H- nasalized in Greek <lagkhánein> 'to fall by lot, obtain by lot or fate, obtain as one's share, have assigned to one'. Frisk notes that this verb has no secure etymology. Nevertheless it must be of inherited IE origin, since a borrowed verb would not show such morphological complexity. Frisk regards the Attic perf. act. <eÃle:kha> and pass. <eÃle:gmai> as neologisms formed after <eÃle:pha> and <eÃle:mmai> from <lambánein> 'to grasp, seize, take', on the grounds that a reduplicated strong perf. <lélogkha> occurs earlier, already in the Odyssey. This observation is reasonable and while on the one hand it nullifies what might be taken as direct evidence for *sl- in the root, on the other hand it explains why some of the forms appear to have contained laryngeals. We can understand the perfects and other long-vowelled forms as analogical creations after the corresponding forms of <lambánein>, whose root is *sleh2gW-. Ionic <lógkhe:> 'share, portion', <eúlogkhos> 'good-fortuned' (Democritus), and <eulog[kh]eîn> 'to enjoy good fortune' (Hesychius), as well as <lélogkha>, obviously do not involve a laryngeal in the root, but it is clear that the nasal has been generalized from the present stem throughout. Goodwin places the verb in Class V without reservation, but I would argue that the strong perfect and /o/-grade derivatives belong to Class II; this is a mixed-class verb (for whatever that might be worth).
The intransitive sense 'fall by lot' of <lagkhánein> is rare, one might say vestigial, in Epic Greek. Typically transitive is the usage of the aforementioned perfect (Od. 11:304): "ti:mè:n dè lelógkhasin îsa theoîsin", 'but they [Castor and Pollux] have obtained honor equally with the gods'. I have assumed that the verb was originally intransitive and impersonal, which implies that an alternative viewpoint developed during the prehistory of Greek. Portions fall by lot or fate to men, and men likewise fall by lot or fate to their portions. The transitive personal usage may have begun with a cognate accusative, the equivalent of "lógkhe:n lakheîn", 'to fall upon one's portion'. Once such a phrase had become established, it would be a simple matter to substitute other nouns for the cognate, hence "ti:mè:n lakheîn", 'to fall upon honor as one's portion, to obtain honor'.
Returning now to <stlocus>, I hypothesize that IE *sleg(^)H- meant essentially 'to be in its assigned place'. It may have had a causative *slog(^)eye/o- 'to put (things) into their assigned places, order, organize'. Fishing with hook and line has been around for many millennia, and anglers need to organize their tackle. Hence one conceivable early usage of IE *slog(^)H-tló-, *sloktló- 'implement for organizing things' may have been 'compartment of a tackle-box', with some sort of collective or plurale tantum denoting the whole box. As technology advanced, naturally the same words would be applied to other compartments and their compartmented artifices. Italic *sloktlos would become Old Latin *stloclos, *stloculus, reinterpreted as a diminutive 'small place' leading to <stlocus>, Classical Latin <locus> 'place'.
DGK