"On the Nature of Man" discusses the humoral theory, but I don't see it describe the observation that drained blood stratifies. The closest I can find in it is "And when men are cut, the blood that flows is at first very hot and very red, and then it flows with more phlegm and bile mixed with it," which is in a section describing how all fluid, when forced from the body, eventually consist of a mixture of humors, in support of the thesis that human nature consists of a mixture of humors rather than a single element.

look the edited answer

RE: attitude toward observation-not so. From Galen's commentary on the very work: "Of matters in dispute based on skill, observation decides some, and reasoning decides others. Things decided by observation, therefore, require a finely discerning observation, and those decided by reasoning require a well-trained argumentation. [...] To know whether ruminants have four stomachs and sheep have one, observation is required, not reasoning."

Oh, hat off to Galen. And my great thanks and +1 to you. I didn't know he spoke about observations before Bacon. But Hippocrate lived about 5 centuries earlier. And I am not sure there were others "admittedly observers" except Galen in these times. Galen could invent that observations of Hippocrate to protect his own thoughts.

... Democritus (460—370 BC) argued that knowledge can be gained from the senses, and Aristotle (384–322 BC) similarly argued for a posteriori knowledge, though the latter also said reason was sufficient to arrive at truth. The Aristotelean method is based on observing phenomena and using reason to discover the form or essence. Note Democritus lived at the same time as Hippocrates.

Hmm. I do not say that they did not use observations. They were not simultaneously deaf and blind. I only say, that the observations were for them something low, not serious, as opposed to the pure thought. That took that from the contemporary understanding of axiom system of geometry. There it (seemingly) worked. And Geometry system of knowledge was the ideal then. And continued to be for 2 thousand years. Aristotle is the best example. He could acknowledge some gain from senses, but it was only words - remember the fly!

You wrote that Hippocrates "would never write about [his observations]" because it wouldn't have been respected.

My point about this is that there were classical philosophers that did respect observations.

Yes. "the observations were for them something low, not serious, as opposed to the pure thought." Sorry for citing, but I didn't want to write it for the third time.

Yes. And he had to look good among this bunch

Everyone needs respect. Excpecially a doctor

Also, you wrote that "giv[ing] post-Bacon philosophy to the scientists of the pre-Bacon times" is an anachronism, which (to me) suggests that using observation when developing at theory is something that only became prevalent with the introduction of the scientific method.

Empiricism existed in various forms well before Bacon.

As for Aristotle, it wasn't just words. What was missing from his method was verification. He would make some observations, then use reason to generalize, but wouldn't then validate the general statement.

Also, we should distinguish (as did the Greeks) "knowledge" from "truth".

Again regarding Aristotle, he failed to account for appearances.

No, not using. But admitting they are using them and using them as an argument. As observations were considered lower than thought, they could be used as an argument against the pure thought, but said earlier.

" he failed to account for appearances." - sorry, don't understand at all

I am not English, as you probably see.

How do you distinguish "knowledge" from "truth"?

I'd say the difference is that knowledge exists in the mind, while truth exists independently of the mind. The various Greek philosophers may have used a different distinction, but they would still distinguish the two. In particular, some would say observation isn't necessary for truth, but it is for knowledge, or (more weakly) that observation could lead to knowledge.

As for Aristotle failing to account for appearances, I was tying in to Simplicius and the Platonic school by equating Plato's challenge to the astronomers to "account for appearances" (σωζειν τα φαινομενα; also translated as "preserve the phenomena" or "save the phenomena") with Aristotle's failure to verify his conclusions by checking them against observations, a major difference between the Aristotelean and scientific methods.

As for you not being a native English speaker, I had suspected, but your English skills are quite good, so I couldn't be sure.

In a way, my overall point is the same point you make about Aristotle. You're using reasoning to come to the conclusion that Hippocrates wouldn't have written about his observations. I only mean to say the matter can only be determined by examining everything we have of Hippocrates' writings (really, it can only be determined by reading everything he wrote, but as that's not possible...).

Hmm. Personally, I don't think there exist any truth independently on mind. Truth is a human term. Even more, I don't think there exist some truths except conditional ones. Independently on us only information exists. But only an intellectual being will say "it is truth" or "it is lie".

I am a simple mathematician and programmer. I don't know any dead language, except some phrases.

Mathematical truths exist independently of a mind, as they rely on a derivation from axioms according to a consistent set of rules, an entirely syntactic endeavor. If truth required a mind, theorem proving programs wouldn't work.

Of course, mathematical truth is quite different from what most people mean by "Truth".

All mathematic is one great conditional truth

That's splitting hairs. Moreover, there are certain statements in logic that are universal.

Every mathematical theory says: if this set of axioms is truth, these theorems are truth, too. Mathematics has many such theories, but not a single unconditional truth.

Tautalogies are unconditional.

No, no! Mathematics so only gives us a tool. When in life we could for some situation prove by experiment, that axioms of some theory are in some situation truth, automatically the whole tree of theorems are truth for us.

Except there's no "when in life" when it comes to math. Tautologies aren't axioms.

Tautalogy, sorry, comrade, is the base rule of the every theory of mathematical logic :-)

What do you mean by "base rule"? That term seems to better describe axioms that tautologies.

Of course, they are not axioms. They are trivial mathematical theories of one sentence each.

base rule only in the sense, that every set of logical axioms starts with a=>a. At least every I remember

Tautologies are statements that are true in all possible interpretations in all possible worlds (in the logical sense of "worlds").

They are unconditionally true.

Simply without them any theory becomes unconstructive. Unusable. I admit that maybe in future there will be some quasitheories without tautology, but I am not SO good in logic to understand the consequences of it

"Aristotle's failure to verify his conclusions by checking them against observations" - that is what I was talking about - if somebody had observation that didn't sit with known truth, woe on him! So it worked in Greece. So it works today, too, but not openly, So, new knowledges come more often.

In any case, we've strayed from my original statement: truth doesn't require a mind. More precisely, a logical statement doesn't require a mind for it to be true (even conditionally true).

Now, knowledge whether a given statement is true or false is another matter.

It's been interesting, but it's well past my bedtime.