Caesar eos in eo loco quo tum ESSENT suum adventum expectare jussit. Caesar ordered them to await his arrival in that place where they then were.
Where did you get this from? The correct quote from De Bello Gallico seems to be:
“Qui cum eum in itinere convenissent seque ad pedes proiecissent suppliciterque locuti flentes pacem petissent, atque eos in eo loco quo tum essent suum adventum expectare iussisset, paruerunt”
To me it seems that the “essent” here is subjunctive by attraction, as a subordinate clause within a clause whose main verb is also in the subjunctive, and with which it forms part of the same idea. See here.
Caesar eos in eo loco quo tum essent suum adventum expectare iussit.
Presumably, this is a simplified version of what Caesar actually wrote.
essent is not “subjunctive by attraction.” There’s no subjunctive verb in a main clause here.
essent is subjunctive because it’s a subordinate clause in indirect speech–it’s part of what Caesar said, not a factual detail supplied by the narrator.
In direct speech, Caesar would have said this, or rather sent ahead this message: Expectate meum adventum In eo loco quo nunc estis.
If Caesar [or someone else] had written Caesar eos in eo loco quo tum erant suum adventum expectare iussit, it would be nonsense: it would mean that Caesar told them in the place where they then were to wait for his arrival–but he would be already there. (Or perhaps it could mean that he told them to stay where they were until he returned after going somewhere else.)
Contrast:
Allen & Greenough sec. 580:
In Indirect Discourse the main clause of a Declaratory Sentence is put in the Infinitive with Subject Accusative. All subordinate clauses take the Subjunctive:—
“sciō mē paene incrēdibilem rem pollicērī ” (B. C. 3.86) , I know that I am promising an almost incredible thing. [Direct: polliceor .]
“nōn arbitror tē ita sentīre ” (Fam. 10.26.2) , I do not suppose that you feel thus. [Direct: sentīs .]
spērō mē līberātum [esse] dē metū; (Tusc. 2.67), I trust I have been freed from fear. [Direct: līberātus sum .]
“[dīcit] esse nōn nūllōs quōrum auctōritās plūrimum valeat ” (B. G. 1.17) , he says there are some, whose influence most prevails. [Direct: sunt nōn nūllī … valet.]
“ nisi iūrāsset, scelus sē factūrum [esse] arbitrābātur ” (Verr. 2.1.123) , he thought he should incur guilt, unless he should take the oath. [Direct: nisi iūrāverō , faciam .]
http://www.perseus.tufts.edu/hopper/text?doc=AG+580&fromdoc=Perseus%3Atext%3A1999.04.0001
with Allen & Greenough sec. 583:
A Subordinate Clause merely explanatory, or containing statements which are regarded as true independently of the quotation, takes the Indicative:—
quis neget haec omnia quae vidēmus deōrum potestāte administrārī; (Cat. 3.21), who can deny that all these things we see are ruled by the power of the gods?
cûius ingeniō putābat ea quae gesserat posse celebrārī; (Arch. 20), by whose genius he thought that those deeds which he had done could be celebrated. [Here the fact expressed by quae gesserat , though not explanatory, is felt to be true without regard to the quotation: quae gessisset would mean, what Marius claimed to have done.]
[*] Note.–Such a clause in the indicative is not regarded as a part of the Indirect Discourse; but it often depends merely upon the feeling of the writer whether he shall use the Indicative or the Subjunctive (cf. §§ 591-593).
http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A1999.04.0001%3Asmythp%3D583
Hylander: you are right of course, it was indirect discourse and not attraction. My mistake.
When I first read the sentence I did think at the time that it might be a subordinate clause in indirect discourse but the sentence for some reason did not look like indirect discourse. I should have gone with my first instinct.
This is the first time I have head of “subjunctive by attraction”. I have several textbooks and none of them mention this topic.
iussit implies speech here.