It may be useful to summarize what we have so far. Cleanthes has presented an argument with the following characteristics:
1. It recognizes an analogy between the world and a machine.
2. On the basis of this analogy, an analogical inference is made on the basic principle of "Like effects have like causes" to the conclusion that, as a machine is designed by a mind, so too the world is designed by a mind.
3. The mind in question must be a human-like mind, since that is the only mind we know.
4. Cleanthes regards one of the strengths of the argument to be that it is a posteriori, i.e., experience-based.
5. He puts it forward as the sole theological argument possible.
6. He claims that it is a sufficient foundation for religion.
Philo originally began by arguing that the analogy is too weak to yield any analogical inference; he is unable to maintain this, and Cleanthes outmaneuvers him in argument. Demea, however, having insisted that we make a sharp distinction between the being and nature of God, argues that the inference has a conclusion that puts too great an emphasis on the similarity of God to a human mind. Philo begins to recognize that the distinction between being and nature sheds important light on the argument; he failed to maintain the argument that the inference to existence was impossible, but that leaves open the question of the nature of what you have inferred to exist.
Thus in Part VI he begins a striking line of argument by proposing alternative analogies. Each of these alternative analogies is intended to explicate the fundamental fact of the order of the world. These analogies are usually understood to be an attack on (2) in the above scheme, but I think it is clear that they are really an attack on (3). Cleanthes's analogical inference has the following structure:
(D) machine : human mind :: world : divine mind
The first alternative hypothesis Philo proposes is:
(WS) animal body : animal soul :: world : world soul
Cleanthes concedes that he had not thought of this, but his first thought in reply is that on such reasoning one might as well say the world is like a vegetable as an animal. A fatal reply. After some discussion of Cleanthes's other suggestions in response to (WS), Philo in Part VII proposes two other analogies:
(G) animal body : generation :: world : something analogous to animal generation
(V) plant body : vegetation :: world : something analogous to vegetation
'Vegetation' here is understood to be the process whereby plants come to exist. You can begin to see what Philo is doing. The question is how to explain the order of the world. Cleanthes has put forward an argument in which we take how a part of the world (machine) comes to be ordered, and infer from that how the whole world comes to be ordered. So Philo naturally begins looking at how other parts of the world come to be ordered (animal bodies, plant bodies).
Demea replies that these are wild and arbitrary hypotheses, and asks what date Philo could have for such conclusions. And Philo replies that this is precisely his point, and the point he has insisted on all along: our experience is so imperfect that it gives us no "probable conjecture" into the order of the whole universe. But if we must make a supposition, we have to do so with the widest possible experience and, given that no other theological argument is allowed, we can have no other rule to guide us by similarity. Thus we have four principles of order in our own little part of the universe: reason (which gives us D), instinct (which gives us WS), generation (which gives us G), and vegetation (which gives us V). Despite what one might think, Philo is not offering each of these analogies as an absolute alternative, as if one has to choose one. Rather, his point is that all these are similar in being principles of order, and are the causes of similar effects. And, what is more, we have no reason to think that they are the only possible principles of order. (Philo is thinking of principles of order in other parts of the universe, but the idea is the same if we take other principles of order in our part of it, e.g., crystallization.) And if we press any one of them too hard, and try to draw too much out of the analogy, we get a radically different cosmogony. D gives us a world that is produced by reason; WS a world produced by instinct; G a world reproduced like an animal; V a world reproduced like a plant.
Cleanthes cannot say that instinct, generation, and vegetation are themselves designed processes without either begging the question or rejecting the a posteriori character of the argument. Precisely what is to be explained is the order of the world; if the argument is to be a posteriori, Cleanthes must consider all the candidates of experience that fit the basic structure of his argument, and not privilege design from the beginning. He can get out of this by assuming that every form of order is a form of design. But if he does he has made the argument a priori!
What is more (and this, I think, is an interesting twist), if Philo choose to think the world was hatched form an egg, Cleanthes has little room to insist that he recognize the world's hatching from an egg as itself caused by a designing mind. For when Philo asked what the designer of the designer was, Cleanthes replied that it didn't matter: it was sufficient to know the cause of the world, and you didn't have to know the cause of the cause of the world to know it. Philo suggests that he can make use of the same defense. If he chooses to infer that the world was hatched from an egg, he doesn't have to give any cause of its generation in this way. If, however, Cleanthes does insist that Philo take the next step and infer that the generation of the world was designed, since there is no other argument allowed by Cleanthes, Philo can with equal right insist that Cleanthes take the next step and infer that the designer was reproduced by animal generation. Equal right? Greater right: because whereas we don't have in our experience much acquaintance with rational minds designing animal generation, we have a great deal of acquaintance with animal generations that result in rational minds, namely, our own reproduction.
Cleanthes responds to this very briefly at the end of the Part VII. I present his answer because its significance is easily missed:
I must confess, Philo, replied Cleanthes, that of all men living, the task which you have undertaken, of raising doubts and objections, suits you best, and seems, in a manner, natural and unavoidable to you. So great is your fertility of invention, that I am not ashamed to acknowledge myself unable, on a sudden, to solve regularly such out-of-the-way difficulties as you incessantly start upon me: though I clearly see, in general, their fallacy and error. And I question not, but you are yourself, at present, in the same case, and have not the solution so ready as the objection: while you must be sensible, that common sense and reason are entirely against you; and that such whimsies as you have delivered, may puzzle, but never can convince us.
People usually write off Cleanthes too early in the Dialogues. Hume, however, did not write Cleanthes as a stupid character. On a superficial reading, Cleanthes' response might seem to be just a stubborn dismissal. It is not. Cleanthes is still pushing on what he already knows to be Philo's weak spot. Philo is a skeptic; he puts a greater emphasis on the natural force of reasoning than on speculative conclusions. With an interlocutor like Philo, Cleanthes does not have to answer every objection, because Philo himself cannot say that every objection should be conceded. Rather, on his own principles we should go (although not dogmatically) with the inference that has the greatest natural force. By claiming that Philo's objections "may puzzle, but never can convince us," and that Philo's claim, that it is more plausible that the universe hatched like an egg than was designed by a mind, is against common sense and reason, he is laying emphasis on this point. It is not stubbornness; it is a sign of just how clever Cleanthes is. Cleanthes' best move at this point is not to budge and not to get distracted by trying to put forward an objection to every alternative speculation Philo puts forward. That's to play the game on Philo's own ground, and Cleanthes has simply not yet been put into a position in which he must play the game on any ground other than his own. It is still the case that Philo's own principles favor Cleanthes' conclusion; at least, Cleanthes can still plausibly claim that they do.
Philo has a response to this. But to get into it would bring us to Part VIII, so we'll just have to wait.