Editing Sampi (section)

==Origins==

Throughout the 19th and 20th centuries, many authors have assumed that sampi was essentially a historical continuation of the archaic letter ''[[san (letter)|san]]'' ({{lang|grc|Ϻ}}), the M-shaped alternative of ''sigma'' (Σ) that formed part of the Greek alphabet when it was originally adopted from [[Phoenician alphabet|Phoenician]]. Archaic san stood in an alphabetic position between ''[[Pi (letter)|pi]]'' (Π) and ''[[Koppa (letter)|koppa]]'' (Ϙ). It dropped out of use in favour of ''sigma'' in most dialects by the 7th century BC, but was retained in place of the latter in a number of local alphabets until the 5th century BC.<ref name="jeffery33">Jeffery, ''Local Scripts'', p. 33.</ref> It is generally agreed to be derived from Phoenician ''[[tsade (letter)|tsade]]''.

The hypothetical identification between san and sampi is based on a number of considerations. One is the similarity of the sounds represented by both. San represented either simple {{IPA|[s]}} or some other, divergent phonetic realization of the common Greek {{IPA|/s/}} phoneme. Suggestions for its original sound value have included {{IPA|[ts]}},<ref name="woodard"/> {{IPA|[z]}},<ref name="jeffery33"/> and {{IPA|[ʃ]}}.<ref name="woodthorpe"/> The second reason for the assumption is the systematicity in the development of the letter inventory: there were three archaic letters that dropped out of use in alphabetic writing (''[[digamma|digamma/wau]]'', ''[[koppa (letter)|koppa]]'', and san), and three extra-alphabetic letters were adopted for the Milesian numeral system, two of them obviously identical with the archaic digamma and koppa; hence, it is easy to assume that the third in the set had the same history. Objections to this account have been related to the fact that sampi did not assume the same position san had had, and to the lack of any obvious relation between the shapes of the two letters and the lack of any intermediate forms linking the two uses.

Among older authorities, Gardthausen<ref>Gardthausen, ''Griechische Palaeographie'', p.369.</ref> and Thompson<ref name="thompson7"/> took the identity between san and sampi for granted. Foat, in a skeptical reassessment of the evidence, came to the conclusion that it was a plausible hypothesis but unprovable.<ref name="foat3"/><ref name="foat2"/> The discussion has continued until the present, while a steady trickle of new archaeological discoveries regarding the relative dating of the various events involved (i.e. the original emergence of the alphabet, the loss of archaic san, the emergence of alphabetic sampi, and the emergence of the numeral system) have continued to affect the data base on which it is founded.

A part of the discussion about the identity of san and sampi has revolved around a difficult and probably corrupted piece of philological commentary by an anonymous [[scholium|scholiast]], which has been debated ever since [[Joseph Justus Scaliger]] drew attention to it in the mid-17th century. Scaliger's discussion also contains the first known attestation of the name "san pi" (sampi) in the western literature, and the first attempt at explaining it. The passage in question is a ''scholion'' on two rare words occurring in the comedies of [[Aristophanes]], ''koppatias'' ({{lang|grc|κοππατίας}}) and ''samphoras'' ({{lang|grc|σαμφόρας}}). Both were names for certain breeds of horses, and both were evidently named after the letter used as a [[Livestock branding|branding]] mark on each: "koppa" and "san" respectively. After explaining this, the anonymous scholiast adds a digression that appears to be meant to further explain the name and function of "san", drawing some kind of link between it and the numeral sign of 900. However, what exactly was meant here is obscure now, because the text was evidently corrupted during transmission and the actual symbols cited in it were probably exchanged. The following is the passage in the reading provided by a modern edition, with problematic words marked:

{{verse translation|lang=grc
| κοππατίας ἵππους ἐκάλουν οἷς ἐγκεχάρακτο τὸ κ[?] στοιχεῖον, ὡς σαμφόρας τοὺς ἐγκεχαραγμένους τὸ σ. τὸ γὰρ σ[?] κατὰ[?] τὸ ϻ[?] χαρασσόμενον ϻὰν[?] ἔλεγον. αἱ δὲ χαράξεις αὗται καὶ μέχρι τοῦ νῦν σῴζονται ἐπὶ τοῖς ἵπποις. συνεζευγμένου δὲ τοῦ κ[?] καὶ σ[?] τὸ σχῆμα τοῦ ἐνακόσιοι[?] ἀριθμοῦ δύναται νοεῖσθαι, οὗ προηγεῖται τὸ κόππα[?]· καὶ παρὰ γραμματικοῖς οὕτω διδάσκεται, καὶ καλεῖται κόππα ἐνενήκοντα.<ref name="holwerda"/>
| "Koppatias" were called horses that were branded with the letter κ[?], just as "Samphoras" were those branded with σ[?]. For the "σ"[?] written like [or: together with?] ϻ[?] was called san[?]. These brandings can still be found on horses today. And from a "κ"[?] joined together with "σ"[?], one can see how the number sign for 900[?] is derived, which is preceded by ''koppa''[?]. This is also taught by the grammarians, and the "90" is called "koppa".
}}

There is no agreement on what was originally meant by this passage.<ref name="soldati"/> While Scaliger in the 17th century believed that the scholiast spoke of ''san'' as a synonym for ''sigma'', and meant to describe the (modern) shape of sampi (ϡ) as being composed of an inverted [[lunate sigma]] and a π, the modern editor D. Holwerda believes the scholiast spoke of the actual M-shaped san and expressed a belief that modern sampi was related to it.

An alternative hypothesis to that of the historical identity between san and sampi is that Ionian sampi may have been a loan from the neighbouring [[Anatolian languages|Anatolian]] language [[Carian]], which formed the local [[substrate (linguistics)|substrate]] in the Ionian colonies of Asia Minor. This hypothesis is mentioned by Jeffery<ref name="jeffery"/> and has been supported more recently by Genzardi<ref name="genzardi"/> 
Brixhe<ref name="brixhe1982"/> suggested that sampi could be related to the Carian letter 25 "<span style="background-color: white;">[[File:Greek Phi 03.svg|x16px|Θ]]</span>", transcribed as ''ś''. This would fit in with the "plausible, but not provable" hypothesis that the root contained in the Carian-Greek names spelled with sampi, "Πανυασσις" and "Οασσασσις", is identical with a root *''uś-/waś-'' identified elsewhere in Carian, which contains the Carian ''ś'' sound spelled with <span style="background-color: white;">[[File:Greek Phi 03.svg|x16px|Θ]]</span>.<ref name="adiego"/> Adiego follows this with the hypothesis that both the Carian letter and sampi could ultimately go back to Greek Ζ (<span style="background-color: white;">[[File:Greek Zeta archaic.svg|x16px|Ζ]]</span>). Like the san–sampi hypothesis, the Carian hypothesis remains an open and controversial issue, especially since the knowledge of Carian itself is still fragmentary and developing.<ref name="wachter"/>

While the origin of sampi continues to be debated, the identity between the alphabetic Ionian sampi ({{IPA|/ss/}}) and the numeral for 900 has rarely been in doubt, although in the older literature it was sometimes mentioned only tentatively,<ref name="thompson7">Thompson, ''Handbook'', p.7.</ref> An isolated position was expressed in the early 20th century by Jannaris, who – without mentioning the alphabetic use of Ionian {{IPA|/ss/}} – proposed that the shape of numeric sampi was derived from a juxtaposition of three "T"s, i.e. 3×300=900. (He also rejected the historical identity of the other two numerals, [[stigma (letter)|stigma]] (6) and [[koppa (letter)|koppa]] (90), with their apparent alphabetic predecessors.)<ref name="jannaris"/> Today, the link between alphabetic and numeral sampi is universally accepted.