SECTION A: INTRODUCTION AND BACKGROUND 1. Framework
3. Methodology
4.3 Use in Achilles’ Text
In Ebbe Vilborg’s introduction to his edition, he has a specific discussion of words containing -ττ- and -σσ-. He explains how he dealt with cases where different branches of the textual tradition differed. Variation between -ττ- and -σσ- (where one form exists in one branch of the manuscript tradition and the other in the other for the same token) is unfortunately rather great. This is especially the case with the lemma for “sea” or
“ocean”, θάλαττα/θάλασσα. Vilborg resisted the urge to generalise one way or the other (as had sometimes been done by earlier editors; Gaselee, for example, generalised θάλασσα throughout). But Vilborg considered it “in point of principle…wrong to write σσ where all our MSS give ττ.”177
My statistics will mainly focus on the number of lemmas and the number of separate tokens containing either the -ττ- or -σσ- form. Based on Vilborg’s edition (run through Watt’s Concordance programme), I found 243 individual tokens containing either -ττ- or σσ- in Achilles’ text. These tokens belong to 52 different lemmas. 11 of these, however, are not relevant to my discussion as they are words that always contain the given spelling and are not known to vary between dialects. These examples include proper nouns such as Ἀττικὸν, Νάρκισσος, Ὀδυσσεὺς and other rare or dialect-specific words (φοινίσσω, μειλίσσω, πατάσσω, (περι)πτύσσω, τυφλώττω, σφάττω, τέττιξ, and ψιττακός). This left me with a total of 223 tokens of the variant in 41 different lemmas.
177 Vilborg 1955: lxxxvi (emphasis original). Gaselee’s edition is the Loeb edition (1917).
62 28 are always spelled with ττ-, 8 always spelled with -σσ- and 5 alternate between the two spellings. The table below shows this data for the 41 relevant lemmas.
Table 4-1 Number of -ττ- vs -σσ- Forms in Achilles Tatius
-ττ- % -σσ- % Total
Total lemmas 28 + 5* 80 8 + 5* 32 41
Total tokens 122 55 101 45 223
*tokens found with both forms
The general pattern shows that Achilles alternates between -ττ- and -σσ-. On the whole, he has a bias towards -ττ- forms. For both lemmas and tokens, -ττ- has the higher number, although while the difference for lemmas is very great (a ratio of 80:32), the difference in total number of tokens is smaller (a ratio of 55:45). Part of the reason for this discrepancy is that the lemma meaning “sea” (θάλασσα) occurs 87 times in the text and is more often spelled with -σσ- than -ττ-.
As I have already cautioned, the fact that we do not have Achilles’ original text must be kept in mind and some variation may be attributable to the manuscript-writers’ own intervention. Because of this, I have gone through Vilborg’s edition and searched for all examples of -ττ-/-σσ- where the manuscript tradition varies (i.e. where one family or some manuscripts show -ττ- and others -σσ- for the same token). The table below shows the statistics with all uncertain tokens removed. The remaining instances are those which have either -ττ- or -σσ- consistently in all manuscripts. As one can see, although the results are somewhat different, the basic trend (more instances off -ττ- as opposed to -σσ-) continues, the ratio becoming 59:41 instead of 55:45.
Table 4-2 Number of -ττ- vs -σσ- Forms Excluding MSS Variation
-ττ- % -σσ- % Total
Total tokens 122 55 101 45 223
Total excluding inconsistent forms 101 59 70 41 171
The conclusion from these statistics is that Achilles Tatius is neither a pure Atticist nor a non-Atticist with regards to the -ττ-/-σσ- marker. He has a slight preference for -ττ- which seems to indicate a desire to Atticise but either he does not feel the need to do so consistently or he is not very good at it and sometimes forgets to replace -ττ- with -σσ-.
63 It is interesting that while for most lemmas he is at least consistent for all instances of that word, 28 of which are always spelled with -ττ- and only 8 with -σσ-, there are 5 in which he is inconsistent even within a single lemma. The consistent lemmas could point him holding that not all -σσ- words needed Atticising with -ττ-, and he may have been under the impression that some of the words for which he kept -σσ- were never spelled with -ττ- in Classical Attic. But for the words where he alternates, this explanation cannot apply.
The table below shows the distribution for the lemmas which vary within Achilles’ text.
The figures in brackets exclude the cases which vary between manuscripts and are less reliable.
Table 4-3 Lemmas with Spelling Variation within Achilles’ Text 178
Lemma -ττ- % -σσ- %
φυλάττω/σσω 7 (6) 78 (86) 2 (1) 22 (14)
γλῶττα/σσα 12 (10) 80 (77) 3 (3) 20 (23) περιττόν/σσόν 3 (3) 75 (75) 1 (1) 25 (25) θάλαττα/σσα 16 (9) 18 (15) 71 (50) 82 (85) τέτταρες/τέσσαρες 5 (4) 29 (44) 12 (5) 71 (56)
The first three words follow Achilles’ general pattern in which -ττ- takes precedence over σσ-. The last two, however, represent anomalies in which the -σσ- spelling is more prevalent. As mentioned earlier, Vilborg found it difficult to accept earlier assumptions that all anomalies were errors and refused to generalise where all manuscripts concurred on a particular form. For the lemma θάλασσα, if I exclude tokens that do vary between the manuscripts there is still a ratio of 15:85. There are at least 9 instances where -ττ- is present in all the manuscripts (along with 50 where -σσ- is invariant) and I suspect that Vilborg is right in assuming that, in the original text, Achilles must have had some alternation between the two forms.
4.3.1 Papyri
As discussed, the variation in the manuscript tradition is something of a problem for my analysis of Achilles’ language. Vilborg’s strict rules in compiling his edition ameliorate
178 In all tables of statistics, unless stated otherwise, figures in brackets represent the number of tokens which are invariant in the manuscripts and therefore likely to be original.
64 this somewhat, but it is also useful to examine what is found in the papyrus fragments which were composed much closer to the date of the original text than the manuscripts.
I found the following:
In the Robinson-Cologne Papyrus (Π4) which is the longest papyrus fragment of Achilles, there are 6 tokens of words containing either -ττ- or -σσ-. Of them, 5 have -ττ- and 1 has -σσ-:
(3.18.5) ΔΕΔΙΤ[ΤΟΜΕΝΟΣ]
(3.20.2) ΘΑΛΑΤΤΗΙ (3.22.6) ΘΑΤΤΟΝ (3.23.5) ΘΑΛΑΤΤΑ (3.25.4) [ΟΡ]ΥΤΤΕΙ (3.20.6) ΤΕ[Σ]ΣΑΡΩΝ.
For δεδιτ[τόμενος], the end of the word from the second -τ- onwards is missing, but the presence of the first -τ- makes -ττ- likely. In τε[σ]σάρων, the first -σ- is missing due to a lacuna in the papyrus. The second -σ- is not entirely clear, but is probably correct, and this reading agrees with all manuscripts. The papyrus forms agree with that given in Vilborg’s edition in all cases except for θαλάττηι and θάλαττα where not all manuscripts agree. The papyrus agrees with the majority (family α and codex F but family β has -σσ- in both cases). This might suggest that, for these instances, the -σσ- form in β might have been a later corruption, but I can only suggest this for these two instances of the θάλασσα word. There are plenty of other instances in which α and F do have -σσ-, either in agreement or disagreement with β.
In the other papyrus fragments, passages containing extant -ττ- or -σσ- are surprisingly rare. In Π1, there is a token of ΕΦΥΛΑΣΣΟΝ (at 2.8.1), with -σσ- in disagreement with Vilborg and all the manuscripts. ΕΝ[ΑΛΛΑΣ]ΣΕΙ also occurs (2.9.1), which is the form that Vilborg uses, although here he chose to use the Π1 form in favour of that found in all the manuscripts, διαλλάσσει. The middle part of the word is missing in the papyrus (as a result of the end of the column having broken off) but the second -σ-, which starts on a new line, is clear. Considering that the manuscripts also have a form in -σσ-, that reading seems reasonable. In Π7, the word ΘΑΛΑΣΣΑ appears once (at 4.14.4). This form agrees with Vilborg and all the manuscripts, providing evidence for early cases of the σσ- spelling of θάλαττα. In sum, it is hard to make any absolute conclusions from
65 the papyrus evidence for -ττ- versus -σσ- because the examples are so few. In some places the papyri agree with some or all the manuscripts, in others they differ. What is significant is that they do not show exclusive use of -ττ- or -σσ-. This substantiates the assumption that variation existed from the earliest version of the text.
4.3.2 A Note on Names and Other Proper Nouns
For the most part, proper nouns are not relevant to the question of Atticism. There are a number of personal and place names that one would not expect to be affected by the variant such as the word for the adjective for something that is Attic, which appears once in the text as Ἀττικόν (2.2.3).
The name of the epic hero, Ὀδυσσεύς (which occurs once in the novel at 2.23.3), is almost exclusively spelled with a -σσ- in literature. Interestingly, it does occur in early Attic inscriptions, especially on early vases, as Ὀλυττεύς, Ὀλυσσεύς and Ὀλυσεύς but the spelling with -δ- unanimously takes -σσ-.179 A full search of the TLG shows only variant spellings with a single -σ- (found even in Homer) and very few examples of Ὀλυσσεύς, Οὐλισές and Οὐλιξεύς (cf. Latin Ulixes, Ulysses) and there are no examples with -(τ)τ-. Achilles’ spelling, therefore, is unremarkable for Greek literature.
According to Threatte, the etymology of the name is unclear, but it is very likely of a non-Greek origin.180
A far more interesting example is the adjective for someone or something from Thessaly, which is typically spelled Θεσσαλός (-ή, -όν) in most dialects, including Thessalian, but is regularly given as Θετταλός (-ή, -όν) in Attic inscriptions and literature.181 In Achilles’ text, it appears 4 times in total, always with the Attic -ττ- spelling (at 5.17.5; 5.22.2; 5.26.12 and 6.16.5). What is especially interesting is the context in which it appears. The word is said twice by Leucippe and twice by Melite of Leucippe. On all 4 occasions in which it occurs, it is used of Leucippe’s fake identity as a Thessalian woman. It may be that Achilles has simply preferred the Attic form as he does elsewhere with -ττ-, but it is possible that by having Leucippe say Θετταλὴ τὸ
179 Threatte 1980: 484, 540
180 Threatte 1980: 484
181 Threatte 1980: 328; 450-451; The Attic Orators exclusively use -ττ- which appears 64 times in their works.
66 γένος (“I am a Thettalian by race”), Achilles is hinting at her deception. It is unlikely that a true Thessalian would use -ττ- for -σσ-.
Finally, the name Melite, though spelled with a single -τ- in Vilborg’s edition, should also be considered here. Vilborg uses the form Μελίτη throughout as it is the spelling found consistently in manuscript family β, but family α regularly has Μελίττη instead.182 Vilborg’s decision for the single -τ- is backed up by the occurrence of ΜΕΛΙΤΗ twice in papyrus fragment Π3 (a third potential token in Π3 has a lacuna right where the Τ or TT would be, and it is unclear). Whether Achilles used double or single -τ- in his original text does not matter so much as that all evidence points to him avoiding the non-Attic form of the same name, Μέλισσα. Threatte identifies Μελίτη as a common form at Athens.183 The Lexicon of Greek Personal Names in Attica shows 6 instances of Μελίτη, 23 of Μελίττα and only 2 of Μέλισσα.184 For this name, then, whether with a single or double -τ-, Achilles also shows an Attic-leaning preference.