A Pyrrhonist's researches do not end in discovery; nor yet do they conclude that discovery is impossible. For they do not terminate at all: the researches continue (PH 1.1, 4), and the researcher finds himself in a condition of ἐποχή (PH 1.7). Έποχή is defined as ‘a standstill of the intellect, as a result of which we neither deny nor affirm anything’ (PH 1.10). The Sceptical investigator neither asserts nor denies, neither believes nor disbelieves.

Έποχή is characteristically produced by argument - indeed, one of the most refreshing features of the Pyrrhonist tracts of Sextus Empiricus is that they are stuffed full of argumentation. When a philosopher offers us an argument, he normally implies that, if we accept the premisses, we ought to accept the conclusion. It is thus natural to suppose that a Pyrrhonist's arguments similarly imply an intellectual ought: ‘Consider these premisses’, the Sceptic urges, ‘and you will see that you should suspend judgement’. A few Pyrrhonian passages do indeed contain such an intellectual ought; but those passages are, I think, misleading. Sextus usually says, not ‘you should suspend judgement’, but ‘you will (or: must) suspend judgement’.

I continue my observations on neglected readings in Horatian manuscripts, begun in these Proceedings, N.S. 15 (1969) and N.S. 17 (1971). I remind readers that my present subject concerns solely the differences between manuscript readings. In that respect my purpose differs from that of P. Maas in SIFC 27–28 (1956) 227–8. While I would not at all object to walking sulle orme di Paul Maas, my opinions on these matters crystallised before he published his paper, and an appraisal of the worth of manuscript readings is a different exercise from emending a corrupt text. In some respects it is easier, in others it is not. The importance of Bentley's work for textual criticism can be understood only by those who have familiarized themselves with both kinds of criticism. Likewise the inadequacy of current editions of Horace will not be brought home to scholars until they have made themselves so aware. A firm basis of what is acceptable in the transmitted text is required before we can hope to achieve agreement on the many places where no manuscript reading will satisfy critical inspection.

15-16 φεύγομεν δ᾿ ἀλώμενοι
ἄλλην ἀπ᾿ ἄλληϲ ἐξορίζοντεϲ πόλιν

The children of Heracles are roaming from city to city in search of refuge. But every city which at first admits them soon yields to the threats of their pursuer Eurystheus and sends them on their way. ‘We wander in exile,’ says Iolaus, ‘from city to city, ἐξορίζοντεϲ. But ἐξορίζειν is a transitive verb, ‘send beyond the frontier, banish’ (LSJ), and so cannot, according to its normal use, have the Heraclidae as its subject. Normal use is illustrated by 257 ϲὺ δ᾿ ἐξόριζε κἆιτ᾿ ἐκεῖθεν ἄξομεν (‘you send them beyond the frontier and we shall take them from there’). LSJ create a special intransitive meaning for the present passage (‘c. acc. loci only … pass from one [city] to another’), just as they do for ὁρίζειν in a passage which Elmsley and others have compared with ours, Med. 433 (‘pass between or through’), though the verb there has its ordinary meaning, as both Pearson (here) and Page (ad loc.) show.

A paradox is like a pun. It is also like a Delphic oracle. For in all three cases, we escape puzzlement, or spoil the joke, when we interpret, when we follow the tracks of the words and disentangle their meaning. So paradoxes are about words - either about the relation between one word and another, or about the relation between words and the world; and the punch of the paradox is delivered by its verbal content. Thus it is characteristic of a good paradox that its verbal content is vicious: paradoxes are very often self-referential, such as ‘Please ignore this notice’.

Paradoxes may be classified according to two main types. Firstly, there are the innocuous paradoxes which tell - or point the way to - a surprising truth. The Socratic Paradoxes, for example, are paradoxical because to say ‘No-one does wrong willingly’ is to contradict the phenomena. But deeper reflection upon Socrates' moral psychology and his account of the good life, might make us concede the truth of his dictum. Certainly it is Socrates' view that we all hold beliefs that entail his thesis. Similarly Heracleitus tells the truth that we cannot step into the same (in all respects) river twice; although if we concede that the waters may flow without damaging the identity of the river, what he says is false. Thus ordinary paradoxes tend to have two faces - their initial, paradoxical one, where they appear false, and their truth, apparent upon reflection.

This is not the age of the amateur. There is every inducement for the specialist to remain a specialist, and if he strays beyond the limits of his particular field, he will be mistrusted, and perhaps rightly. But there are times when he has no choice but to do so, and to risk the consequences; there are some fields which are so untrodden that the amateur will find himself in good company, or perhaps in no company at all. And there are times when the classicist, the Old Testament scholar, and certainly the Egyptologist will need to remind themselves that the world of the eastern Mediterranean was never a large one, and that communications in the first millennium B.C. were not so rudimentary. A month's journey could have covered much ground, or sea: and some people had the courage to travel far longer than this.

Herodotus (5.115) records how, shortly after the outbreak of the Ionian revolt, the Persians launched a counter-attack, retook Sardis in Lydia, and turned south across the river Maeander into the thickly-wooded areas beyond. It was no time to be a defender. Two disastrous engagements are recorded, together with some Herodotean numbers of dead, but the few survivors succeeded in laying a night ambush for the invading Persians, and cut them to pieces. All three Persian commanders perished. As it happened, it was no time to be a Persian.

1. 151-7 qualiter expressum ventis per nubila fulmen
aetheris impulsi sonitu mundique fragore
emicuit rupitque diem populosque paventes
terruit obliqua praestringens lumina flamma:
in sua templa furit, nullaque exire vetante
materia magnamque cadens magnamque revertens
dat stragem late sparsosque recolligit ignes.

Of the rival interpretations of templa in 155, ‘temples’ and ‘quarters of the sky’ Housman espoused the latter; but he did not explain sua, which Weise explained as ‘loca aut regna, in quibus oritur et quasi habitat et regnat fulmen’ (‘the quarter of the sky whence it came’ Haskins). Such in fact seems the only possible interpretation, given the chosen sense of templa, despite Duff's ‘it rushes to its appointed quarter of the sky’ or Bourgery's ‘elle fond sur les lieux qu'elle consacre.’ But why should Lucan tell us this (Weise's ‘habitat et regnat’ is only his own embroidery), and so spoil his sequel? How a Roman reader would understand templa should be obvious to anyone who has read Pease's note on Cic. Div. 1.19 ipse suos quondam tumulos aut templa petivit/et Capitolinis iniecit sedibus ignis.