P

edit with form

Character
Customary name:	pi

Variants and attestation

Transliteration	Sinistroverse		Dextroverse
	Glyph	Number	Glyph	Number
P		8		9
P2		29		3
P3		1		0
P4		2		0
P5		1		0
P6		1		0

Commentary

Pi is one of the letters (together with Lambda and Upsilon) which serve as a basis for distinguishing the Magrè and Sanzeno alphabets (see Script): The variants , , , with an angle (sometimes rounded), often termed "Venetoid" in the TIR, are attributed to the Magrè alphabet, while with a bar on top extending against writing direction is used in the Sanzeno alphabet.

While the schibboleth letters mentioned above do usually appear in the expected forms and are grouped together appropriately in numerous inscriptions from both alphabetical contexts, the letter form ← / → (commonly Magrè Lambda) does crop up in Sanzeno context, where it is not only incongruous, but in some cases must clearly be read /p/ instead of /l/. The identification of ← / → is problematical in the following circumstances:

• Inscriptions in which ← appears together not with Magrè Pi , but with Sanzeno Lambda and/or Sanzeno Pi ←.
• Inscriptions with isolated ← (not accompanied by any other variant of Pi or Lambda), whose linguistic content (e.g. phonetical plausibility, but most prominently comparison with independently attested material) suggests a reading /p/.

Both types of cases are so far only known from Sanzeno context, wherefore we may add a third, not purely epigraphically motivated condition:

• Inscriptions from (epigraphical or geographical) Sanzeno context with isolated ← which cannot be definitely demonstrated to write /l/.

← appears combined only with Sanzeno Lambda on CE-1.3 (twice) and SZ-22.1, and is consequently considered to represent Pi turned against writing direction in both cases. Both inscriptions duly have Sanzeno Upsilon , though the one from geographically intermediate Cembra also features word-internal punctuation. With the bar of Pi extending in writing direction, the distinction between Pi, Lambda and Upsilon is still sustained, and this system is actually the one in use in the Lugano alphabet. However, both inscriptions display clearly Raetic features in both writing (←, ←) and content.

It is not clear, how and why the non-Venetic characteristics of the Sanzeno alphabet arose, but a letter variant ← could easily be explained as an intermediate form between and which might reasonably be expected to crop up sporadically as a marginal archaism. However, in SZ-87 ← appears combined with Sanzeno Pi . The two letters occur, separated by only one letter, within the same word ?ipis, probably an individual name in the genitive. Here, ← would a priori be considered a case of alphabetical interference, Magrè Lambda in a Sanzeno inscription – maybe written by a person proficient in both alphabets. But a reading lipis precludes the comparison of the form with pipe on BZ-11 (not autopsied). A reading pipis can be argued by grouping SZ-87 with the three inscriptions which have the triple combination of ←, ← and ←, being one Sanzeno Lambda, one Sanzeno Pi, and one letter of doubtful ascription. These are WE-3, SZ-15.1 and SZ-30. In WE-3, the dubious ← occurs in the name piθamnuale, which is attested several times – in Magrè context with clear Pi, but in one other instance in Sanzeno context also with ← (see BZ-9 below). WE-3 has three incontestably correct Sanzeno Lambdas, and one Sanzeno Pi in the name laspa, which is also attested on SZ-1.1 and, incidentally, SZ-15.1. SZ-15.1 has one Sanzeno Lambda in laspa, two Sanzeno Pis in laspa and ?ipie, and two instances of ← in ?ipie and ka?aśunu. ?ipie may be compared with ?ipis in SZ-87 (the corresponding nominative?); on the segmentation of and possible comparanda for ka?aśunu see the inscription page. SZ-30, the only dextroverse one of the inscriptions discussed here, has Sanzeno Lambda twice in well attested forms, Sanzeno Pi in the anlaut of an isolated name (?) pumis, and twice anlauting → in obscure words.

Isolated ← is read Pi based on content-related arguments in the inscriptions BZ-9 and possibly NO-13. On the name piθame in BZ-9 see above (WE-3). For the question of whether ← is intended in NO-13, see the inscription page; if it is the case, perisna is well attested also. ← also appears on its own in NO-11, but without parallel attestation no decision can be made between the readings liri and piri.

Regarding solely its form, the letter ← might in Sanzeno context be considered either an influence from the Magrè alphabet (when alongside Sanzeno Pi; in SZ-87), or a remnant form of Pi (when alongside Sanzeno Lambda or isolated; in BZ-9, CE-1.3 and SZ-22.1). While it is not absurd to presume two different origins and consequently values of the form, neither of these explanations accounts for the cases where it appears alongside both regular Sanzeno letter forms. The postulation of simple writing mistakes is not advisable. For WE-3, this explanation might pass, seeing as ← occurs only in the completely inverted second line as opposed to ← in the first, in combination with the fact that Sigma also appears turned in different directions. But Sigma is actually the very letter whose orientation is known to occasionally vary within one inscription – letters with bars are generally written with surprising uniformity. Writing ← instead of ←, which is the only letter with a bar extending "backwards" from a straight hasta, may seem an easy mistake to make, but there appears to have been no problem with the equally counterintuitive ←. In SZ-15.1, ← and → occur almost successively; in SZ-30, the writer would have switched from "wrong" to "correct" and back again, and that only in that part of the inscription which is left to us. Such crude mistakes stand in contrast to the noticably neat execution of the inscriptions discussed. While downright blunders are improbable, the redundantly distinctive retrograde bar of ← may have been considered not obligatory by some writers, making the form interchangeable with its archaic variant ←, though in this case a higher frequency of occurrence of ← should be expected.

Note that among twelve instances of ←, two words appear twice: the names piθam(n)e and ?ipie. The former might etymologically belong with other individual names in piθ-, one of which is piθiave; the second one of the forms in question on SZ-30 might also belong with this group. Yet it is unlikely that ← vs. ← reflects a phonetic reality – while more than the usual number of characters for labial stops seem to have been used by the Raeti, and the problematic case of NO-13 might connect ← with Φ5 s, the name piθam(n)e is attested with regular Magrè Pi in Serso (see Φ).

In light of the fact that in none of the cases discussed above, a reading /l/ is preferable to a reading /p/, ← in Sanzeno context is consistently defined as Pi in TIR and transliterated accordingly, unless there is a reason to do otherwise. This includes fragmentary inscriptions which have not been mentioned (VN-2, VN-3, SZ-32) and a great number of inscriptoids from Sanzeno (most prominently the upi-inscriptions). The only exception so far is the la-group, which appears to have alternating with .

Bibliography

Kluge & Salomon 2015	Sindy Kluge, Corinna Salomon, "Ausgewählte Funde aus Dercolo im Kontext der rätischen Inschriften", Wissenschaftliches Jahrbuch der Tiroler Landesmuseen 8 (2015), 80–95.
Markey 2006	Thomas L. Markey, "Early Celticity in Slovenia and at Rhaetic Magrè (Schio)", Linguistica 46 (2006), 145–171.