Nunlist ————— Greek, Roman, and Byzantine Studies 60 (2020) 124–138 Article copyright held by the author(s) and made available under the Creative Commons Attribution License CC-BY https://creativecommons.org/licenses/by/4.0/ Nicanor’s System of Punctuation René Nünlist HE ANCIENT GRAMMARIAN Nicanor of Alexandria (second century A.D.) is the author of an elaborate system of punctuation. Owing to its complexity, the system was and is a potential source of confusion among readers. The present contribution attempts to clarify the picture by giving a concise account as a set of rules, each illustrated with examples. The goal is to expound the system as such, not to discuss what might have led Nicanor to develop it.1 1. There are two principle sources for Nicanor’s system of punctuation: (i) the comprehensive description in schol. Dion. Thr. 26.4–28.8 Hilgard, which provides a short definition and an example for each punctuation mark; (ii) the relevant scholia to the Iliad that discuss individual passages.2 The two sources 1 This paper owes much to the only modern attempt to give a compre- hensive description of Nicanor’s system: L. Friedländer, Nicanoris περὶ Ἰλιακῆς στιγμῆς reliquiae emendatiores2 (Berlin 1857). To revisit the question is never- theless justified for two reasons: (i) Friedländer’s account is not well suited for quick reference because it is fairly long (well over 100 pages) and written in Latin. (ii) Friedländer spends comparatively little time on describing Ni- canor’s system and rationale as such. Instead there is a curious tendency to focus on passages where, to Friedländer’s mind, Nicanor got it wrong; cf. D. Blank, “Remarks on Nicanor, the Stoics and the Ancient Theory of Punctua- tion,” Glotta 61 (1983) 48–67, at 48. 2 The text of schol. Dion. Thr. is reprinted in the Appendix below, together with a selection of the Homeric scholia that form the backbone of this paper. For summaries see C. Wendel, “Nikanor (27),” RE 17 (1936) 274–277, at 276–277 (in German), and Blank, Glotta 61 (1983) 49–51 (in English). Useful as they are, they both omit the examples and, in the case of the ἀνυπόκριτος ὑποστιγµή (see §3.2), fail to point out the errors that are explained in n.18 below. T RENÉ NÜNLIST 125 ————— Greek, Roman, and Byzantine Studies 60 (2020) 124–138 largely concur and complete each other, with the following picture emerging: the system consists of eight marks in total, five στιγµαί (see §2), two ὑποστιγµαί (§3) and one διαστολή (§4). The eight marks differ with a view to function, delivery, and typogra- phy.3 2. The hardest punctuation mark is the στιγµή. It is roughly equivalent to a full stop and put after sentences that are con- sidered complete (αὐτοτελής). There are five variants (in de- scending order of ‘hardness’): 2.1 The τελεία στιγµή is put when the next sentence is con- nected asyndetically (i.e. without a connecting particle). Examples:4 (a) µάντι κακῶν οὔ πώ ποτέ µοι τὸ κρήγυον εἶπας. | αἰεί τοι τὰ κάκ᾿ ἐστὶ φίλα φρεσὶ µαντεύεσθαι (Il. 1.106–107). (b) οὐκ ἀγαθὸν πολυκοιρανίη. εἷς κοίρανος ἔστω (Il. 2.204, cf. schol. A Il. 2.204–205 Nic.). (c) ἀλλὰ σὺ … περίσχεο παιδὸς ἑῆος. | ἔλθουσ᾿ Οὔλυµπόνδε Δία λίσαι κτλ. (Il. 1.393–394, cf. schol. A Il. 1.393a Nic.). Vocatives (with or without imperative phrase) are regularly rounded off by a τελεία στιγµή.5 To Nicanor’s mind, Homer is fond of asyndeton (which leads to a τελεία στιγµή) in transitions (µεταβάσεις).6 In delivery, the τελεία στιγµή is indicated by four χρόνοι of 3 The present account focuses on the function of the punctuation marks and briefly touches upon the topic of delivery. For the (less than clear) typo- graphical details, see Wendel, RE 17 (1936) 276–277. 4 The examples in the section on στιγµαί follow a tripartite structure: (a) is the example given in schol. Dion. Thr. (usually supplemented in order to make it easier to understand); (b) is an example that Nicanor’s note expressly attributes to this specific category; (c) is an example that can be shown to fall into this specific category even though Nicanor’s note indicates only the gen- eral category στιγµή (see §2.6). 5 Schol. A Il. 1.59a Nic., schol. A Il. 1.106a Nic. (i.e., after µάντι κακῶν in example (a), without schol. Dion. Thr. taking note of it), schol. A Il. 2.56a Nic., etc. 6 Schol. A Il. 11.150 Nic. 126 NICANOR’S SYSTEM OF PUNCTUATION ————— Greek, Roman, and Byzantine Studies 60 (2020) 124–138 silence (schol. A Il. 2.23b Nic.; see Appendix). 2.2 The ὑποτελεία στιγµή is put when the next sentence is connected by means of a particle (e.g. δέ, γάρ, ἀλλά, αὐτάρ). Examples: (a) πολλὰς δ᾿ ἰφθίµους ψυχὰς Ἄϊδι προΐαψεν | ἡρώων. αὐτοὺς δὲ ἑλώρια τεῦχε κύνεσσιν (Il. 1.3–4). (b) εἴ περ γάρ κ᾿ ἐθέλῃσιν Ὀλύµπιος ἀστεροπητής, | ἐξ ἑδέων στυφελίξαι. ὁ γὰρ πολὺ φέρτατός ἐστιν (Il. 1.580–581, cf. schol. A Il. 1.580–583a Nic.).7 (c1) ἀλλ᾿ ἤτοι ἔπεσιν µὲν ὀνείδισον ὡς ἔσεταί περ. | ὧδε γὰρ ἐξερέω (Il. 1.211–212, cf. schol. A Il. 1.211–212a1 Nic.). (c2) τετρήχει δ᾿ ἀγορή. ὑπὸ δὲ στεναχίζετο γαῖα (Il. 2.95, cf. schol. A Il. 2.95–96 Nic.). The ὑποτελεία στιγµή is indicated by three χρόνοι of silence.8 2.3 The πρώτη ἄνω στιγµή is put after a sentence that contains a µέν, ἤ, or οὐ that in turn is taken up by δέ, ἤ (ἦ), or ἀλλά in the next sentence. Examples: (a) αἴδεσθεν µὲν ἀνήνασθαι· δεῖσαν δ᾿ ὑποδέχθαι (Il. 7.93). (b) οἳ µὲν κήρυσσον (Nic.: ἐκήρυσσον vulg.)· τοὶ δ᾿ ἠγείροντο µάλ᾿ ὦκα (Il. 2.52, cf. schol. A Il. 2.52a Nic.; see Appendix).9 (c) ἤ ῥ᾿ αὖτις πόλεµός τε κακὸς καὶ φύλοπις αἰνὴ | ἔσσεται· ἦ φιλότητα µετ᾿ ἀµφοτέροισι τίθησι | Ζεύς (Il. 4.82–84 [with Nicanor’s accentua- tion], cf. schol. A Il. 4.82–83a1 Nic.). 7 This is the only note that expressly mentions the ὑποτελεία στιγµή (and this particular solution, which takes στυφελίξαι to be an optative, is only one of three suggestions for this passage). At first sight, this finding is surprising, especially when held against the substantial number of notes that expressly mention the τελεία στιγµή (twenty on the first three books of the Iliad alone). A possible answer might be that asyndeton (i.e. τελεία) is the exception and connection by means of particles (i.e. ὑποτελεία) the rule. No other στιγµή will have been as frequent in Nicanor’s text as the ὑποτελεία, which therefore counts, so to speak, as the default. 8 This is an inference from the fact that the τελεία στιγµή has four χρόνοι and the πρώτη ἄνω στιγµή two. It is confirmed by schol. h Il. 2.877b (printed in Erbse’s Testimonienapparat and the Appendix below), see Friedländer 119. 9 This is the only example found in the extant notes on the Iliad. RENÉ NÜNLIST 127 ————— Greek, Roman, and Byzantine Studies 60 (2020) 124–138 The πρώτη ἄνω στιγµή is indicated by two χρόνοι of silence (schol. A Il. 2.52a Nic.; see Appendix). 2.4 The δευτέρα ἄνω στιγµή is put between sentences that are connected by means of καί. Examples: (a) καί ῥα πάροιθ᾿ αὐτοῖο καθέζετο· καὶ λάβε γούνων (Il. 1.500). (b) Ζεῦ ἄνα, δὸς τίσασθαι (…) | δῖον Ἀλέξανδρον· καὶ ἐµῇς ὑπὸ χερσὶ δάµασσον (Il. 3.351–352, cf. schol. A Il. 3.352a Nic.).10 (c) καί κε τὸ βουλοίµην· καί κεν πολὺ κέρδιον ἦεν (Il. 3.41, cf. schol. A Il. 3.41–42 Nic.). The δευτέρα ἄνω στιγµή is indicated by one χρόνος of silence (schol. A Il. 2.131–132 Nic.; see Appendix). 2.5 The τρίτη ἄνω στιγµή is put between sentences that are connected by means of τε. Examples: (a) Κίλλαν τε ζαθέην· Τενέδοιό τε ἶφι ἀνάσσεις (Il. 1.38).11 (b) No example preserved. (c) Γλαῦκος δ᾿ ἔγνω ᾗσιν ἐνὶ φρεσί· γήθησέν τε κτλ. (Il. 16.530, cf. schol. A Il. 16.530–531a Nic.). The τρίτη ἄνω στιγµή is in all likelihood indicated by one χρόνος of silence. 2.6 In practice, Nicanor’s notes, at least in their extant form, regularly fail to specify which of the five στιγµαί is actually meant. Likewise, the instruction to start a new sentence (ἀπ᾿ ἄλλης ἀρχῆς προφέρεσθαι vel sim.) presupposes a preceding στιγµή, just as the instruction to take a clause by itself (καθ᾿ ἑαυτό) presupposes a στιγµή on each end. In all these cases, readers are expected to decide for themselves which type of στιγµή is meant. 10 This is one of only two notes that explicitly mention the δευτέρα ἄνω στιγµή, the other being schol. A Il. 2.131–132 Nic. 11 The example is curious because it comes from what most readers would analyse as a relative clause ( ὃς Χρύσην ἀµφιβέβηκας | Κίλλαν τε ζαθέην κτλ.). Nicanor perhaps took ὅς to be a demonstrative pronoun, as he sometimes does (Friedländer 34). 128 NICANOR’S SYSTEM OF PUNCTUATION ————— Greek, Roman, and Byzantine Studies 60 (2020) 124–138 2.7 Normally speaking, a στιγµή requires the relevant sentence to be complete (αὐτοτελής). Nicanor is, however, prepared to admit some exceptions when the sense requires a στιγµή even though the sentence is not, strictly speaking, complete. These στιγµαί are called ἐν αἰτήµατι.12 3. The two ὑποστιγµαί and the διαστολή (§4) represent softer punctuation marks. They are put after a clause that is incom- plete and are therefore roughly equivalent to a comma.13 3.1 The ἐνυπόκριτος ὑποστιγµή marks the pivot between subordinate clause and main clause (in this sequence, which Nicanor considers an ὀρθὴ περίοδος, cf. §4.1).14 The rule applies to all preceding subordinate clauses alike (temporal, causal, con- ditional, purpose, comparative, relative, etc.).15 Examples:16 (a) ἦµος δ᾿ ἠριγένεια φάνη ῥοδοδάκτυλος Ἠώς, | καὶ τότ᾿ ἔπειτ᾿ ἀνάγοντο κτλ. (Il. 1.477–478). (b1) οὕνεκα δὴ νῦν δῖον Ἀλέξανδρον Μενέλαος | νικήσας ἐθέλει στυγερὴν ἐµὲ οἴκαδ᾿ ἄγεσθαι, | τοὔνεκα δὴ νῦν δεῦρο δολοφρονέουσα παρέστης (Il. 3.403–405, cf. schol. A Il. 3.403–405a1 Nic.; see 12 All four extant examples (schol. A Il. 2.404–407 Nic., schol. A Il. 2.681– 685 Nic., schol. A Il. 11.677b Nic., schol. A Il. 14.317a Nic.) deal with a κοινόν, which normally receives a διαστολή (§4.2). 13 The approximate equation στιγµή ≈ full stop and ὑποστιγµή/διαστολή ≈ comma must be read against the backdrop of the observation that “Ni- canor’s eight-fold system is the elaboration of a two-fold system, distinguish- ing complete and incomplete phrases” (Blank, Glotta 61 [1983] 51). 14 The term ἐνυπόκριτος no doubt means that the relevant clause receives a particular delivery (ὑπόκρισις), for instance, by raising the pitch of the voice from where it descends into the main clause (Friedländer 59). 15 The definition in schol. Dion. Thr. appears to presuppose a neat cor- relation between subordinate and main clauses (ὄφρα – τόφρα, ἦµος – τῆµος, ὅτε – τότε, ἕως – τέως, ὅπου – ἐκεῖ), which may represent something like the ideal case (cf. example (b1) with Nicanor’s note). The total of Nicanor’s notes, however, does not support such a narrow definition; hence the more general description given above in the main text. 16 The paragraph on ὑποστιγµαί can do without examples of the type (c) because Nicanor’s notes provide enough examples of the type (b). RENÉ NÜNLIST 129 ————— Greek, Roman, and Byzantine Studies 60 (2020) 124–138 Appendix).17 (b2) εἰ δέ κε µὴ δώῃσιν, ἐγὼ δέ κεν αὐτὸς ἕλωµαι | ἐλθὼν σὺν πλεόνεσσι (Il. 1.324–325, cf. schol. A Il. 1.324–325 Nic.). When several subordinate clauses precede the main clause they are separated by additional ἐνυπόκριτοι ὑποστιγµαί. Example: (b) ἀλλ᾿ ὅτε δὴ σόλον εἷλε µενεπτόλεµος Πολυποίτης, | ὅσσον τίς τ᾿ ἔρριψε καλαύροπα βουκόλος ἀνήρ, | ἣ δὲ ἑλισσοµένη πέτεται διὰ βοῦς ἀγελαίας, | τόσσον παντὸς ἀγῶνος ὑπέρβαλε (Il. 23.844–847, cf. schol. A Il. 23.844–847 Nic.; see Appendix). The ἐνυπόκριτος ὑποστιγµή is indicated by one χρόνος of silence (schol. A Il. 8.206–207a1 Nic.). This probably applies to the other ὑποστιγµή as well. 3.2 The ἀνυπόκριτος ὑποστιγµή marks the pivot between the end of an insertion (διὰ µέσου, i.e. parenthesis broadly under- stood) and the main clause. Examples: (a) [Il. 3.33–36].18 (b) οἳ δὴ νῦν ἕαται σιγῇ, πόλεµος δὲ πέπαυται, | ἀσπίσι κεκλιµένοι (Il. 3.134–135 [the insertion underscored], cf. schol. A Il. 3.134b Nic.; see Appendix). The punctuation mark before the insertion depends on whether 17 Like all ancient punctuation systems, Nicanor’s lacks a question mark. The fact that line 405 is a question must be put into words (ἠθικῶς πυνθάνεται, schol. T Il. 3.405 Nic., cf. πευστικῶς in schol. A Il. 1.290–291 Nic., etc.). The same holds true mutatis mutandis for the exclamation mark, which modern edi- tors still eschew (indicated by θαυµασµός/θαυµαστικός in schol. A Il. 5.638b Nic., etc.). 18 The example in schol. Dion. Thr. is more confusing than helpful because the author has made two fundamental mistakes. The last word in the διὰ µέσου part (here παρειάς, Il. 3.35) is normally marked by an ἀνυπόκριτος ὐποστιγµή (never an ἐνυπόκριτος ὑποστιγµή). And the word right before the διὰ µέσου part (here βήσσῃς, 3.34) never receives an ἀνυπόκριτος ὑποστιγµή (see below). Moreover, the author wrongly claims that these insertions always fall between the subordinate and the main clause in ὀρθαὶ περίοδοι. In reality, the insertion can ‘interrupt’ all kinds of constructions (see e.g. example (b)). 130 NICANOR’S SYSTEM OF PUNCTUATION ————— Greek, Roman, and Byzantine Studies 60 (2020) 124–138 it rounds off a subordinate or a main clause. In the former case §3.1 applies (ἐνυπόκριτος ὑποστιγµή), in the latter §4.1 (δια- στολή), as in the example. To mark the end of an insertion (διὰ µέσου) appears to be the sole function of the ἀνυπόκριτος ὑποστιγµή.19 It seems likely that Nicanor invented this particular punctuation mark for a syn- tactic phenomenon that he recognised in at least 65 Iliadic passages.20 4. The βραχεῖα διαστολή (henceforth: διαστολή) is Nicanor’s softest punctuation mark.21 As such it is the one whose range of applications is particularly wide. 4.1 The διαστολή marks the pivot between the main clause and the subordinate clause (in this inverted sequence, which Nicanor considers an ἀνεστραµµένη περίοδος, cf. §3.1). The rule applies to all subsequent subordinate clauses alike. Examples: (a) ἦ κεν γηθήσαι Πρίαµος Πριάµοιό τε παῖδες | ἄλλοι τε Τρῶες µέγα κεν κεχαροίατο θυµῷ, | εἰ σφῶϊν τάδε πάντα πυθοίατο µαρναµένοιϊν (Il. 1.255–257). (b) καί νύ κεν εἴρυσσέν τε καὶ ἄσπετον ἤρατο κῦδος, | εἰ µὴ ἄρ᾿ ὀξὺ νόησε Διὸς θυγάτηρ Ἀφροδίτη (Il. 3.373–374, cf. schol. A Il. 3.374a Nic.; see Appendix). When several subordinate clauses follow the main clause they 19 This is a minor simplification because long insertions can receive ad- ditional ἀνυπόκριτοι ὑποστιγµαί in the middle (cf. schol. A Il. 1.234–240a Nic.), cf. the multiple ἐνυπόκριτοι ὑποστιγµαί that connect several subordinate clauses (§3.1). 20 For Nicanor’s treatment of διὰ µέσου in general, see R. Nünlist, “Two Cornerstones of Nicanor’s Syntactic Explanations,” RivFil 147 (2019) 395– 416. In exceptional cases, the insertion (διὰ µέσου) is marked on either side with a διαστολή (Friedländer 79–80, with ref. to schol. A Il. 13.787a Nic.). The reason seems to be that the insertion is very short. Note, however, that the equally short insertion in Il. 11.242 receives the regular punctuation (schol. A Il. 11.242–243a1 Nic.). 21 Given that there is one διαστολή only, the specification βραχεῖα is strictly speaking superfluous, which is why it is frequently omitted. The term in schol. Dion. Thr. is ὑποδιαστολή. RENÉ NÜNLIST 131 ————— Greek, Roman, and Byzantine Studies 60 (2020) 124–138 are separated by additional διαστολαί. Example: (b) ἦ µένετε Τρῶας σχεδὸν ἐλθέµεν, ἔνθα τε νῆες | εἰρύατ᾿ εὔπρυµνοι πολιῆς ἐπὶ θινὶ θαλάσσης, | ὄφρα ἴδητ(ε) κτλ. (Il. 4.247–249, cf. schol. AbT Il. 4.247–248 Nic.). When a subordinate clause is surrounded by a main clause, a διαστολή marks the beginning of the subordinate clause and an ἐνυπόκριτος ὑποστιγµή its end (in accordance with §3.1). Example: (b) σὺ δὲ σῷ µεγαλήτορι θυµῷ | εἴξας, [διαστ.] ἄνδρα φέριστον ὃν ἀθάνατοί περ ἔτισαν, [ἐνυπ. ὑποστ.] | ἠτίµησας (Il. 9.109–111, cf. schol. A Il. 9.110a Nic.; see Appendix).22 The function of the διαστολή described in this paragraph is the only one that has an explicit theoretical basis in the ancient material. What follows is an attempt to deduce further categories from recurrent types of notes and their respective explanations.23 4.2 The διαστολή separates clauses that share a common ele- ment (κοινόν) and are therefore not complete. Examples: ἀµφὶ δ᾿ ἄρ᾿ ὤµοισιν βάλετο ξίφος ἀργυρόηλον | χάλκεον, αὐτὰρ ἔπειτα (sc. ἀµφ᾿ ὤµοισιν βάλετο) σάκος µέγα τε στιβαρόν τε (Il. 3.334–335, cf. schol. A Il. 3.335a1 Nic.; the reader must mentally supply the predicate phrase). ἀλλ᾿ ἄγε Πατρόκλῳ ἔφεπε κρατερώνυχας ἵππους, | αἴ κέν πώς µιν ἕλῃς, δώῃ δέ τοι εὖχος Ἀπόλλων (Il. 16.724–725, cf. schol. A Il. 16.725 Nic.: the διαστολή after ἕλῃς is owed to the fact that, as a κοινόν, the conjunction αἴ κεν governs both subordinate clauses).24 4.3 The διαστολή marks off appositions and comparable forms of epexegesis. 22 Whereas modern editors put a comma after φέριστον, Nicanor wants the διαστολή to come after εἴξας. 23 Therefore, a differentiation between different types of examples (a, b, c) is no longer appropriate. 24 As an alternative, the note considers putting a στιγµή (sc. ὑποτελεία), with δώῃ δέ τοι εὖχος Ἀπόλλων becoming an independent sentence. For Nicanor’s treatment of κοινόν in general, see Nünlist, RivFil 147 (2019) 408–414. 132 NICANOR’S SYSTEM OF PUNCTUATION ————— Greek, Roman, and Byzantine Studies 60 (2020) 124–138 Examples: τῇσι θύρας ὤϊξε Θεανὼ καλλιπάρῃος | Κισσηΐς, ἄλοχος Ἀντήνορος ἱπποδάµοιο (Il. 6.298–299, cf. schol. A Il. 6.299 Nic.). ἥδε δέ οἱ κατὰ θυµὸν ἀρίστη φαίνετο βουλή, | πέµψαι ἐπ᾿ Ἀτρείδῃ Ἀγαµέµνονι οὖλον Ὄνειρον (Il. 2.5–6, cf. schol. A Il. 2.5b Nic.; see Appendix). 4.4 The διαστολή separates words in juxtaposition that must not be taken together. Or, put differently, it indicates which words are to be taken together. Examples: ὃ µὲν εὔχετο πάντ᾿ ἀποδοῦναι, | δήµῳ πιφαύσκων (Il. 18.499–500, cf. schol. A Il. 18.499–500b Nic.: δήµῳ is the indirect object of πιφαύ- σκων, not ἀποδοῦναι). (Zeus) ἀστράψας δέ, µάλα µεγάλ᾿ ἔκτυπε (Il. 17.595, cf. schol. A Il. 17.595 Nic.: µάλα µεγάλ(α) is an intensifier of the thunder, not the lightning). οὗτοι ἄρ᾿ ἡγεµόνες Δαναῶν, ἕλον ἄνδρα ἕκαστος (Il. 16.351, cf. schol. A Il. 16.351a Nic.: to take Δαναῶν with ἄνδρα would create an in- consistency [ἀπεµφαῖνον]). αὐτοὶ δ᾿ (…) | στείοµεν, εἴ κεν πρῶτον ἐρύξοµεν ἀντιάσαντες(,) | δούρατ᾿ ἀνασχόµενοι (Il. 15.296–298, cf. schol. A Il. 15.298 Nic., where he deliberates whether or not to put a διαστολή after ἀντι- άσαντες or, in other words, whether δούρατ᾿ ἀνασχόµενοι goes with the main or the subordinate clause). All examples, except for the last, punctuate the text in a way that might strike modern readers as odd because it impedes the reading flow. Nicanor, however, is keen to make the syntactic construction as transparent as possible. He says so explicitly in the case of the fairly numerous διαστολαί that are put πρὸς τὸ σαφέστερον (or the like).25 25 Schol. A Il. 11.4 Nic., schol. A Il. 11.227a Nic., schol. A Il. 11.397–398 Nic., etc. Friedländer (84) perceives a difference between schol. A Il. 16.65– 70 Nic. and schol. A Il. 16.68a Nic. and concludes that a διαστολὴ πρὸς τὸ σαφέστερον is optional. Such an understanding clashes with the fact that Ni- canor regularly uses the verbal adjective διασταλτέον (incl. Friedländer’s own example) and, even more tellingly, declares such διαστολαί to be ἀναγκαῖαι in schol. A Il. 6.38–41 Nic., as Friedländer himself acknowledges (84 n.2). For an optional διαστολή see e.g. schol. A Il. 13.260a Nic. RENÉ NÜNLIST 133 ————— Greek, Roman, and Byzantine Studies 60 (2020) 124–138 4.5 The διαστολή can be used to segment lists of similar words/expressions, for instance, a chain of epithets. Example: ἔχε δ᾿ αἰγίδα θοῦριν, | δεινήν, ἀµφιδάσειαν, ἀριπρεπέ(α) (Il. 15.308– 309, cf. schol. A Il. 15.308–309 Nic.: here and elsewhere Nicanor is of the opinion that such a punctuation/delivery results in increased emphasis). 5. Even though it is not, strictly speaking, a punctuation mark, it may nevertheless be worth mentioning another lectional sign in Nicanor’s system: the συναφή. It refers to a sublinear mark that combines two to three words which together form one semantic unit. Example: …, ἐπεὶ ἦ καὶ ἐµὸν βέλος ὀξὺ πάροιθεν (Il. 20.437, cf. schol. A Il. 20.437b1 Nic., which considers the two underlined words a unit and glosses them with τὸ ὠξυµµένον κατὰ τὸ ἔµπροσθεν). The alternative term for συναφή is ὑφ᾿ ἕν (schol. A Il. 12.228– 229a Nic.). 6. No fewer than five punctuation marks are indicated by one χρόνος of silence (δευτέρα and τρίτη ἄνω στιγµή, both ὑπο- στιγµαί, and the διαστολή). It is not clear how the listener is to tell the difference between, say, a τρίτη ἄνω στιγµή and a δια- στολή (the two ὑποστιγµαί differ in intonation). Nor is it clear whether Nicanor discussed the issue. APPENDIX Schol. Dion. Thr. 26.4–28.8 Hilgard: ἵνα δὲ µὴ δόξῃ τις ἡµᾶς ἀγνοεῖν καὶ τὴν τοῦ λεχθέντος Νικάνορος διατύπωσιν τὴν περὶ τῶν στιγµῶν, ὧν τὰ ὀνόµατα ἤδη ἡµῖν προείρηται (sc. p.24.16–18 Hilgard), δεῖ ὡς ἐν συντόµῳ ἐνταῦθα µνησθῆναι τῆς τε θέσεως αὐτῶν καὶ τῆς διαφορᾶς τῆς τῷ Νικάνορι εἰρηµένης· (…) Lest anyone believe that we are unaware of said Nicanor’s system of punctuation, the names of which we have mentioned already, it is necessary to give here a brief summary on where to put the marks and how this differs from what Nicanor says. (…) [2.1] ἡ µὲν οὖν τελεία στιγµὴ τίθεται ἐν τῷ µέσῳ τόπῳ τῆς τελευταίας γραµµῆς τοῦ στοιχείου τοῦ τελευταίου ἐν τοῖς ἀσυνδέτοις λόγοις, ὡς ἐπὶ 134 NICANOR’S SYSTEM OF PUNCTUATION ————— Greek, Roman, and Byzantine Studies 60 (2020) 124–138 τοῦ “µάντι κακῶν, οὔ πώ ποτέ µοι τὸ κρήγυον εἶπας” (Il. 1.106) εἰς τὸ “εἶπας,” τουτέστιν εἰς τὸ τελευταῖον αὐτοῦ γράµµα ἡ τελεία τίθεται, τοῦ ἐφεξῆς λόγου (sc. αἰεί τοι τὰ κάκ᾿ ἐστὶ φίλα κτλ.) ἀσυνδέτου ὄντος. The τελεία στιγµή (lit. complete point) is put in the middle area of (i.e. behind) the final stroke of the final letter in asyndetic sentences; for instance, in µάντι κακῶν, οὔ πώ ποτέ µοι τὸ κρήγυον εἶπας, on (i.e. behind) εἶπας, that is, the τελεία is put on its final letter be- cause the subsequent sentence is asyndetic. [2.2] ἡ δὲ ὑποτελεία ὀλίγον ὑποκάτω τοῦ µέσου τόπου τοῦ στοιχείου τοῦ ἐσχάτου, ὅτε ἐπιφέρεται ὁ “δέ” ἢ ἄλλος τις σύνδεσµος τῶν ἰσοδυναµούν- των τῷ “δέ,” λέγω δὲ τὸν “γάρ,” τὸν “ἀλλά,” τὸν “ἀτάρ,” τὸν “αὐτάρ.” ὡς ἐπὶ τοῦ “ἡρώων. αὐτοὺς δὲ ἑλώρια” (Il. 1.4) εἰς τὸ ν τοῦ “ἡρώων” ἡ ὑποτελεία τίθεται διὰ τὸ ἐπιφέρεσθαι τὸν “δέ.” οὕτω γὰρ δοκεῖ τῷ ἡµετέ- ρῳ γραµµατικῷ, φηµὶ δὲ τῷ Ἀπολλωνίῳ, καὶ µάλα εὖ, ὥς γέ µοι δοκεῖ· ἔφη δὲ ἐν τούτῳ παρεωρακέναι τὸν Νικάνορα, τὴν µὲν ἐν τῷ χρόνῳ τῆς σιωπῆς διαφορὰν ὁρισάµενον τῶν πρώτων δύο στιγµῶν, φηµὶ δὲ τῆς τε- λείας καὶ τῆς ὑποτελείας, τὴν δὲ θέσιν καὶ τὸν τόπον τὸν αὐτὸν ταῖς δύο ἀπονέµοντα· οὐ δεῖ οὖν τὴν τελείαν καὶ τὴν ὑποτελείαν ἐν τῷ µέσῳ τοῦ στοιχείου τιθέναι, ὥς γέ φησι Νικάνωρ, ἐπεὶ ποία διαφορὰ ἔσται αὐτῶν; ἀλλὰ τὴν µὲν τελείαν ἐν τῷ µέσῳ, τὴν δὲ ὑποτελείαν ὑποκάτω ὀλίγον τοῦ µέσου τόπου τοῦ τελευταίου στοιχείου. The ὑποτελεία <στιγµή> (lit. less than complete point) slightly below the middle area of (i.e. behind) the final letter when δέ follows (sc. in the next sentence) or some other conjunction with the same force as δέ, namely γάρ, ἀλλά, ἀτάρ, αὐτάρ; for instance, in ἡρώων. αὐτοὺς δὲ ἑλώρια, the ὑποτελεία is put on (i.e. behind) the nu of ἡρώων, owing to the subsequent δέ. For this is how our grammarian, I mean Apollonius (sc. Dyscolus), sees it, and rightly so, as it appears to me. He said that in this respect Nicanor was wrong when, on the one hand, he determined the difference in duration of the silence (i.e. pause) of the first two punctuation marks, I mean the τελεία and the ὑποτελεία, and, on the other, attributed the position (sc. within the sentence) and the same place to both.26 The τελεία and the ὑποτελεία are not to be placed both in the middle of (i.e. behind) the letter, as Nicanor claims; for what will be the difference between them? Instead the τελεία 26 In light of the opposing terms τελεία vs. ὑποτελεία, this is unlikely to be an accurate rendering of what Nicanor actually said on the issue (Friedländer 4). RENÉ NÜNLIST 135 ————— Greek, Roman, and Byzantine Studies 60 (2020) 124–138 in the middle, and the ὑποτελεία slightly below the middle area of (i.e. behind) the final letter. [2.3] ἡ δὲ πρώτη ἄνω τίθεται ἐπάνω τῆς τελευταίας γραµµῆς τοῦ τελευ- ταίου στοιχείου, ὅτε πρόκειται ὁ “µέν” ἢ ὁ “ἤ” ἢ τὸ “οὔ,” ἐπιφέρεται δὲ ὁ “δέ” ἢ ὁ “ἤ” ἢ ὁ “ἀλλά”, ὡς ἐπὶ τοῦ “αἴδεσθεν µὲν ἀνήνασθαι·” (Il. 7.93) εἰς τὸ τελευταῖον ι τοῦ “ἀνήνασθαι” ἡ πρώτη ἄνω τίθεται διὰ τὸ ἐπιφέρεσθαι τὸν “δέ” (sc. δεῖσαν δ᾿ ὑποδέχθαι), τοῦ “µέν” προκειµένου. The πρώτη ἄνω <στιγµή> (lit. first high point) is put above the final stroke of the final letter when µέν or ἤ or οὐ precedes and δέ or ἤ or ἀλλά follows (sc. in the next sentence); for instance, in αἴδεσθεν µὲν ἀνήνασθαι, the πρώτη ἄνω is put on the final iota of ἀνήνασθαι be- cause δέ follows (sc. in the next sentence), with µέν preceding. [2.4] ἡ δὲ δευτέρα ἄνω τίθεται καὶ αὐτὴ ἐπάνω τῆς τελευταίας γραµµῆς τοῦ ἐσχάτου στοιχείου, περιέχεται δὲ ὑπὸ διπλῆς ἔξωθεν, ὅτε ἐπιφέρεται ὁ “καί,” ὡς ἐπὶ τοῦ “καί ῥα πάροιθ᾿ αὐτοῖο καθέζετο· καὶ λάβε γούνων” (Il. 1.500) εἰς τὸ ο τοῦ “καθέζετο” τίθεται ἡ δευτέρα ἄνω, τοῦ “καί” ἐπι- φεροµένου. The δευτέρα ἄνω <στιγµή> (lit. second high point) too is put above the final stroke of the last letter and is embraced by a diplê from outside (i.e. ·>) when καί follows (sc. in the next sentence); for instance, in καί ῥα πάροιθ᾿ αὐτοῖο καθέζετο· καὶ λάβε γούνων, the δευτέρα ἄνω is put on the omicron of καθέζετο because καί follows. [2.5] ἡ δὲ τρίτη ἄνω τίθεται καὶ αὐτὴ ἐπάνω µὲν τῆς τελευταίας γραµµῆς τοῦ τελευταίου στοιχείου, περιέχεται δὲ ὑπὸ διπλῆς ἔσωθεν, ὅτε ἐπι- φέρεται ὁ “τέ,” ὡς ἐπὶ τοῦ “Κίλλαν τε ζαθέην· Τενέδοιό τε ἶφι ἀνάσσεις” (Il. 1.38) εἰς τὸ ν τοῦ “ζαθέην” τίθεται ἡ τρίτη ἄνω, ἐπιφεροµένου τοῦ “τέ” συνδέσµου. The τρίτη ἄνω <στιγµή> (lit. third high point) too is put above the final stroke of the final letter and is embraced by a diplê from inside (i.e. <·) when τε follows (sc. in the next sentence); for instance, in Κίλλαν τε ζαθέην· Τενέδοιό τε ἶφι ἀνάσσεις, the τρίτη ἄνω is put on the nu of ζαθέην because the conjunction τε follows. [3.1] ἡ δὲ ὑποστιγµὴ ἡ ἐνυπόκριτος τίθεται ὑποκάτω µὲν τῆς τελευταίας γραµµῆς τοῦ τελευταίου στοιχείου, ὀλίγον δὲ ἐξωτέρω ἐκ πλαγίου νεύ- ουσα, ἐν ταῖς ὀρθαῖς περιόδοις, τουτέστιν ὅτε πρόκειται τὸ “ὄφρα” ἢ τὸ “ἦµος” ἢ τὸ “ὅτε” ἢ τὸ “ἕως” ἢ τὸ “ὅπου,” ἐπιφέρεται δὲ τὸ “τόφρα,” τὸ “τῆµος,” τὸ “τότε,” τὸ “τέως,” τὸ “ἐκεῖ,” καὶ τὰ ὅµοια, ὡς ἐπὶ τοῦ “ἦµος δ᾿ ἠριγένεια φάνη ῥοδοδάκτυλος Ἠώς” (Il. 1.477 etc.) εἰς τὸ ς τοῦ “Ἠώς” τίθεται ἡ ἐνυπόκριτος, ἐπιφεροµένου τοῦ “τότε” (e.g. καὶ τότ᾿ ἔπειτ᾿ ἀνάγοντο κτλ.). 136 NICANOR’S SYSTEM OF PUNCTUATION ————— Greek, Roman, and Byzantine Studies 60 (2020) 124–138 The ἐνυπόκριτος ὑποστιγµή (lit. point below with special intonation) is put below the final stroke of the final letter, slightly outside displaced sideways (i.e. to the right?27) in ‘straight periods’, that is, when ὄφρα or ἦµος or ὅτε or ἕως or ὅπου precede (sc. in the subordinate clause) and τόφρα, τῆµος, τότε, τέως, ἐκεῖ, or the like follow (sc. in the sub- sequent main clause); for instance, in ἦµος δ᾿ ἠριγένεια φάνη ῥοδοδάκτυλος Ἠώς, the ἐνυπόκριτος is put with (i.e. below) the sigma of Ἠώς because τότε follows (sc. in the main clause). [3.2] ἡ δὲ ὑποστιγµὴ ἡ ἀνυπόκριτος τίθεται καὶ αὐτὴ ὑποκάτω τοῦ τελευταίου γράµµατος ὑπὸ τὴν ἐσχάτην καὶ κατωτάτην γραµµὴν τοῦ στοιχείου, ἐν δὲ ταῖς µεταξὺ πρὸ τῆς ἀνταποδόσεως τῶν ὀρθῶν περιόδων ἀναφωνουµέναις ἑτέραις περιόδοις, ὡς ἐπὶ τοῦ “ὡς δ᾿ ὅτε τίς τε δράκοντα ἰδὼν παλίνορσος ἀπέστη οὔρεος ἐν βήσσῃς” (Il. 3.33–34) ἐνταῦθα τίθεται ἡ ἀνυπόκριτος, ὁµοίως καὶ εἰς τὸ “γυῖα” 3.34) καὶ εἰς τὸ “ἀνεχώρησεν” (3.35)· µεταξὺ γὰρ <πρὸ> τῆς ἀνταποδόσεως ἕτεραι περίοδοι ἐνετέθησαν· εἰς δὲ τὸ “παρειάς” (3.35) ἡ ἐνυπόκριτος· εὐθέως γὰρ ἐπιφέρεται ἡ ἀνταπόδοσις. The ἀνυπόκριτος ὑποστιγµή (lit. point below without special intonation) too is put below the final letter, below the last and lowest stroke of the letter, in the intervening parts of the sentence that are uttered before the main clause of ‘straight periods’; for instance, in ὡς δ᾿ ὅτε τίς τε δράκοντα ἰδὼν παλίνορσος ἀπέστη οὔρεος ἐν βήσσῃς, the ἀνυπόκριτος is put here (sc. below the final letter of βήσσῃς) and likewise with γυῖα and ἀνεχώρησεν. For other parts of the sentence have been inserted before the main clause. An ἐνυπόκριτος <ὑποστιγµή> is put with (i.e. below the final letter of) παρειάς. For the main clause fol- lows immediately upon it.28 [4.1] ἡ δὲ ὑποδιαστολὴ κατὰ πάντα τῶν προλαβουσῶν στιγµῶν ἐνήλ- λακται, ὡς καὶ αὐτῷ τῷ στίγµατι µὴ εἶναι νυγµή τις, ὡς αἱ ἄλλαι, ἀλλὰ τὸν τύπον ἔχειν τοῦ ὀξέος τόνου· τίθεται δὲ καὶ αὐτὴ ὑποκάτω τῆς ἐσχάτης γραµµῆς τοῦ τελευταίου γράµµατος, ὡς ὀξεῖα δέ τις, ὡς προ- είρηται, ἐν ταῖς ἀντεστραµµέναις περιόδοις, ὡς ἐπὶ τοῦ “ἦ κεν γηθήσαι Πρίαµος <Πριάµοιό τε παῖδες> ἄλλοι τε Τρῶες µέγα κεν κεχαροίατο θυµῷ, εἰ σφῶϊν τάδε πάντα πυθοίατο µαρναµένοιιν” (Il. 1.255–257) ἐν τῷ “θυµῷ” τίθεται ἡ ὑποδιαστολή· ἐπιφέρεται γὰρ τὸ “εἰ σφῶϊν” καὶ 27 Neither is the exact position of the mark clear nor how it differs from the confusingly similar position of the ἀνυπόκριτος ὑποστιγµή. 28 For the errors in this section see n.18 above. RENÉ NÜNLIST 137 ————— Greek, Roman, and Byzantine Studies 60 (2020) 124–138 ποιεῖ ἀντεστραµµένην τὴν περίοδον· ἡ γὰρ ὀρθὴ περίοδος ἦν “εἰ σφῶϊν τάδε πάντα πυθοίατο, ἦ κεν γηθήσαι Πρίαµος” (Il. 1.257+255). The ὑποδιαστολή (comma below) is completely different from the previous punctuation marks in that even in its form it is not a dot as the others, but has the shape of an acute accent. It too is put below the last stroke of the final letter, just like an acute , as men- tioned, in ‘inverted periods’. For instance, in ἦ κεν γηθήσαι Πρίαµος <Πριάµοιό τε παῖδες> ἄλλοι τε Τρῶες µέγα κεν κεχαροίατο θυµῷ, εἰ σφῶϊν τάδε πάντα πυθοίατο µαρναµένοιιν, a ὑποδιαστολή is put on (i.e. below the final letter of) θυµῷ because εἰ σφῶϊν (i.e. the sub- ordinate clause) follows and inverts the period. The ‘straight period’ would be εἰ σφῶϊν τάδε πάντα πυθοίατο, ἦ κεν γηθήσαι Πρίαµος. Schol. A Il. 2.23b Nic.: τελεία στιγµὴ κατὰ τὸ τέλος, ἐπεὶ ἀσύνδετος ὁ λόγος. τέσσαρας δὲ αὕτη χρόνους σιωπῆς δύναται. a τελεία στιγµή at the end (sc. of the line) because the sentence is asyndetic. It is equivalent to four χρόνοι of silence. Schol. h Il. 2.877b: ὑποτελεία γὰρ ἔσται, ἥτις τρίχρονος δύναται εἶναι (codd.: τρεῖς χρόνους δύναται Friedländer). There will be a ὑποτελεία (sc. at the end of book 2), which is equiv- alent to three χρόνοι. Schol. A Il. 2.52a Nic.: ἐπὶ τὸ “κήρυσσον” ἄνω πρώτη στιγµή, ἥτις δύο δύναται χρόνους σιωπῆς. an ἄνω πρώτη στιγµή on κήρυσσον, which is equivalent to two χρόνοι of silence. Schol. A Il. 2.131–132 Nic.: ἐπὶ δὲ τὸ “πλάζουσι” (132) θετέον τὴν ἄνω δευτέραν στιγµήν, ἥτις ἕνα χρόνον δύναται σιωπῆς. One must put an ἄνω δευτέρα στιγµή on πλάζουσι, which is equiv- alent to a single χρόνος of silence. Schol. A Il. 3.403–405a1 Nic.: ὑποστικτέον δὲ ἐν ὑποκρίσει εἰς τὸ “ἄγεσθαι” (404), ἐπεὶ τῶν ἀνταποδοτικῶν ἐστι καὶ ταῦτα, “οὕνεκα” (403) — “τοὔνεκα” (405). An ἐνυπόκριτος ὑποστιγµή must be put on ἄγεσθαι, because these too belong to the correlatives, οὕνεκα — τοὔνεκα.29 29 The expression ἐν ὑποκρίσει is a common variant of ἐνυπόκριτος (cf. schol. A Il. 9.110a Nic., also in this Appendix), but can mean other things as well, for instance, “in dissimulation” (i.e. ironically, e.g. schol. Eur. Or. 488). 138 NICANOR’S SYSTEM OF PUNCTUATION ————— Greek, Roman, and Byzantine Studies 60 (2020) 124–138 Schol. A Il. 23.844–847 Nic.: ὑποστικτέον “Πολυποίτης” (844), “ἀνήρ” (845), “ἀγελαίας” (846)· τῇ γὰρ προτέρᾳ προτάσει ἑτέρα ἐπενήνεκται. An <ἐνυπόκριτος> ὑποστιγµή must be put on Πολυποίτης, ἀνήρ, ἀγελαίας. For another subordinate clause follows upon the first.30 Schol. A Il. 3.134b Nic.: διὰ µέσου τὸ ἡµιστίχιον “πόλεµος δὲ πέπαυ- ται”· διόπερ διαστέλλοµεν ἐπὶ τὸ “σιγῇ,” στίζοµεν δὲ ἀνυποκρίτως ἐπὶ τὸ “πέπαυται.” The half-verse πόλεµος δὲ πέπαυται is an insertion (lit. in the middle); therefore we put a comma on σιγῇ and an ἀνυπόκριτος ὑποστιγµή on πέπαυται. Schol. A Il. 3.374a Nic.: πάλιν διαστολὴ πρὸ τούτου· ἀνέστραπται γὰρ ἡ περίοδος. ὀρθὴ δ᾿ ἂν ἐγένετο οὕτως· “εἰ µὴ ἄρ᾿ ὀξὺ νόησε, καί νύ κεν εἴρυσ<σ>έν τε.” again a comma before this line because the period is inverted. A straight would have read thus: εἰ µὴ ἄρ᾿ ὀξὺ νόησε, καί νύ κεν εἴρυσ<σ>έν τε. Schol. A Il. 9.110a Nic.: ὁ λόγος αἱρεῖ διαστέλλειν ἐπὶ τὸ “εἴξας,” ὑποστίζειν δὲ ἐν ὑποκρίσει ἐπὶ τὸ “ἔτισαν.” It seems good to put a comma on εἴξας, an ἐνυπόκριτος ὑποστιγµή on ἔτισαν. Schol. A Il. 2.5b Nic.: βραχὺ διασταλτέον κατὰ τὸ τέλος· µέρος γὰρ κατὰ (cod.: καὶ Friedländer) τὸ ἐπιφερόµενον τῆς περιόδου. One must put a short comma (sc. and not a ‘harder’ punctuation mark) at the end (sc. of the line) because what follows equally be- longs to the period. December, 2019 Institut für Altertumskunde Albertus-Magnus-Pl. D-50923 Cologne, Germany rene.nuenlist@uni-koeln.de 30 There can be no doubt that the ἐνυπόκριτος is meant even though the scholion does not say so explicitly.