Think about it this way: in ethics, or at least in virtue ethics, we say that a virtuous person does certain things (does the right thing, acts in accordance with the mean, acts as the virtuous agent would, or what have you) and we also say that a virtuous person is a sort of way (has certain character traits, is generous, is brave, and so forth). To give an account of the ethically virtuous person, we need to give some sort of account of what that person does and how that person is. And it’s not at all clear, and is often explicitly denied, that giving an account of what the person does is sufficient to understanding how the person is. To fully understand the nature of the ethically virtuous person, we need to give an account of her capacities for action (and her actions themselves) and her character. I think the same holds for the account of the knower. The knower does certain things (has certain capacities or powers that are actualized; for Plato, the knower will be able to survive dialectic, will be able to give an account of the aitia, and so forth) but also is a certain way. She has certain character traits, of both the intellectual and, for Plato, the ethical, variety. And these character traits are essential features of being a knower. So, just as in the case of the ethically virtuous person, here too in order to fully understand the nature of the knower, we need to give an account both of the intellectual capacities of the knower (the capacities one has in virtue of having knowledge) and the traits that characterize the knower.

So far so good, right? Well. Sort of. There’s something that I need to argue for before I can just jump in to this chapter. What is that? I’m making an assumption here (one that is probably glaringly obvious to most folks) — I’m assuming that giving a full account of the knower will involve more than simply giving an account of the nature of knowledge. I’m assuming that there are important character traits that are essential features of the knower and that we can’t understand them just by giving an account of the nature of knowledge. Understanding the nature of knowledge is not sufficient to understand the nature of the knower. And that’s not an assumption I just get to make. I need to argue that this is something that we can see in Plato.

Here’s a more general way to think about it: I need to argue that the relationship between the dunamis and the possessor of that dunamis is a complex, rather than simple, relationship. What do I mean by that? Most broadly, the question is asking whether we can fully characterize the nature of the possessor of a dunamis simply in virtue of the nature of the dunamis itself, or whether the possessor of the dunamis possesses other things (what sorts of things? Character traits, other abilities, who knows…) that are not explained in virtue of the possession of that dunamis and thus, giving an account of a dunamis will not be sufficient for giving an adequate account of the possessor of that dunamis. Whew. That’s a mouthful. And thoroughly unclear, I think.

Let’s us the dunamis of rhetoric as an example. What is the relationship between rhetoric and the rhetorician? Is it a simple one — will giving an account of the dunamis of rhetoric fully characterize the rhetorician (qua rhetorician)? Or is the relationship more complex — will we be missing something in our characterization of the rhetorician if we only give an account of rhetoric? Let’s say that we can give a full account of the nature of rhetoric. It’s a dunamis and we are able to fully describe the powers and capacities that it imparts to the possessor of that dunamis. The rhetorician, in virtue of being a rhetorician will be able to do x, y, and z. If we are able to fully characterize the dunamis of rhetoric, will we thereby have fully characterized the nature of the rhetorician qua rhetorician (e.g. is a rhetorician simply someone who can do x, y, and z)? Or, even if we are able to fully articulate the dunamis of rhetoric, may we not need to still describe something more in order to give a full account of the nature of the rhetorician?

So.
Simple relationship between a dunamis and the possessor of that dunamis: one can fully understand the nature of the possessor of a dunamis by understanding the nature of the dunamis itself.
Complex relationship between a dunamis and the possessor of that dunamis: one cannot fully understand the nature of the possessor of a dunamis by understanding the nature of the dunamis itself. (Because the possessor of the dunamis has certain features that aren’t explained by virtue of the nature of the dunamis itself.)
The general question: Is there a simple or complex relationship between a dunamis and the possessor of that dunamis in Plato’s discussions of dunameis?

At this point, though, I’m wavering between two different ways to approach this question.

1) I can start by looking at the more particular question of the relationship between the dunamis of knowledge and the knower. This will involve focusing a great deal on Republic six (where we can see Plato describe the characteristics of the knower but in an importantly different way than he characterized the dunamis of knowledge in Republic five). From there (having shown that the relationship, in the case of knowledge, is complex) I can then move on to the more general question and see whether what holds for knowledge holds for other dunameis and if not, why not.

2) I can start by thinking about the topic more generally. What is the relationship between the dunamis and the possessor of that dunamis? Is it simple? Complex? Are there some instances in which it is simple and others in which it is complex? If it is this last case (which, at the moment, I think it is) then how can we explain why it is that in some cases the relationship is simple and in others it is complex? (And, of course, where does the dunamis of knowledge fit in to this schema?)

I’m leaning towards the second approach at this point (although I’ve already written up an introduction to the issue wherein I say I’m going to opt for the first approach so clearly my thoughts at the moment are rather fluid). What might be very useful at this point, though, is just to look at three or four different dunameis that Plato talks about and see how he characterizes the dunamis itself and the possessor of that dunamis. That’s going to be pretty crucial for either approach here. I’ve already done this a bit with rhetoric and will try to write something up about it soon. I wonder what other dunameis might be useful to look at here.

The Stoics hold that knowledge is very difficult to attain and yet it is attainable. Because knowledge is attainable, nature must provide us with the means to acquire knowledge. It does this by providing us with clear and distinct impressions. Stoic epistemology is based on the doctrine of clear and distinct impressions and so the skeptics focused their debate on that aspect of their epistemology.

The Stoic Position

Impressions

Both humans and animals have impressions, though the impressions of humans (qua rational) are rational impressions.

Rational impressions (RI) have propositional content (we should interpret this notion of ‘impression’ in the way that we would interpret it in the statement ‘the impression which one gets, if one looks at the evidence, is that…’)

RI’s present themselves to the mind and the mind either accepts them or refuses to accept them.
* to accept them is to have the belief that the proposition is true
* to refuse to accept them is to suspend judgment regarding the veracity of the proposition

The Stoics held that the σοφος will never have any false beliefs. This is possible because humans are able to distinguish mere impressions from καταληψεις without fail.
* These καταληψεις provide an ample foundation for what we need to know
* By only accepting καταληψεις we may not accept many true impressions, but we will never accept a false impression.

This account presupposes two things:
1) there are καταληψεις
2) the mind can distinguish between καταληψεις and other impressions
To understand how the Stoics can make these assumptions, we must look at the Stoic account of rational impressions.

All RI’s, even the most primitive, involve concepts and cognitions. Part of what an object does is make the mind conceptualize it in a particular way.

“RI’s and in particular cognitive impressions do presuppose concepts, but these arise from more primitive impressions that do not presuppose these concepts, and ultimately from sense impressions that do not presuppose any concepts whatsoever but that are not rational either” (154).

So we have a non-ad hoc account of the development to RI’s.

There is more to an RI than the propositional content. To have an RI is to think of a certain proposition in a certain way. The kind of impression that it is depends on 1) the propositional content and 2) how the propositional content is thought.

Impressions:
* are impressions of an object
* consist of a thought concerning the object
* the thought involves a conceptualization of the object
* this thought need not be entirely conceptual
* the thought is characterized by the proposition AND the way the proposition is thought
* the way the proposition is thought depends on the way the constituents of the thought are represented in the thought

Cognitive Impressions (καταληψεις)

Impressions have a causal history and in the course of this history all sorts of things can go wrong. HOWEVER, we can imagine that nature has made it such that under normal conditions the impression we receive is true. So impressions with the right causal history are guaranteed (presumably by nature) to be true. We can call these normal impressions.

There are at least two kinds of normal impressions:
1) when the NI is not produced or caused by the object of the impression itself. I might have the NI that 2+2=4 because I have a proof that 2+2=4.
2) When the NI is produced by the object of the impression. This only happens in the case of perception. Looking at an apple gives me the impression of an apple; this impression is caused by the object of the impression. We can call these perceptual impressions.

At SE 7.424 we are given five conditions which must be met for a visual impression to be cognitive:
1) conditions on the sense organ
2) conditions on the object of vision
3) conditions on how the object is placed
4) conditions on how the impression comes about
5) conditions on the state of mind of the perceiver

The Stoics appear to think that cognitive impressions are perceptual. There’s a worry with this idea, though.
* They also clearly assume that there are non-perceptual cognitions (in cases where we have a proof of a theorem, for example).
* But even non-perceptual cognitions involve impressions.
* So it seems natural to assume that the impressions involved in cognitions are cognitive, even if they’re non-perceptual.
* But if there are non-perceptual cognitions, then there should be non-perceptual cognitive impressions.
* Besides, cognitive impressions are supposed to be a criteria of truth such that they guarantee the truth of all other impressions that the σοφος accepts. If we restrict cognitive impressions to those garnered from perception, it is hard to see how this could act as a sufficient foundation for the σοφος.

How can we respond to this worry?
First, we can see that the Stoics hold that all features of objects are perceptible. We can learn to see that something is good just like we can learn to see that something is a man. Perception, then, provides a much broader basis than we would initially assume.

Also, all other impressions can be accepted as true to the extent that their truth is guaranteed by the truth of perceptual impressions.
“It seems that the Stoics take on the view that only perceptual impressions are cognitive in their own right. Thus other impressions can be called cognitive only to the extent that they have a cognitive content which depends on the cognitive content of impressions which are cognitive in their own right” (159).

SO, there are two types of impressions:
1) self-evident impressions that are cognitive in a narrow sense (perceptual impressions (?))
2) evident impressions that are cognitive in a wider sense. (impressions that depend on perceptual impressions (?))

Perceptual impressions represent the objects the way they do because the objects are that way. The represented features of the impression are due to the object and not some abnormal condition.

So a perceptual impression in no way misrepresents its objects, under normal conditions. We can tell clearly what its visual features are; these impressions are evident.

Evidence is an objective feature of impressions in much the same way that having a clear view of something is an objective fact, not a subjective feeling.

Impressions are true because their propositional content is true; it doesn’t depend on the way the propositional content is thought.

Evidence is a feature of impressions which does depend on the way a proposition is represented by thought. The proposition “this is blue” can be evident when thought because I see the blue ink while non-evident when I hear from someone else that the ink is blue.

“What makes a thought or impression evident is that it is already part of the representation of the subject of the proposition that the predicate should be true of it and that the representation of the subject is entirely due to the subject itself. Thus evidence is not what makes an impression or proposition true, but an evident proposition cannot but have a true proposition for its content and hence be true itself” (161).

A cognitive impression will be evident in that it represents all features of the object appropriate for the kind of object it is.

CI’s differ by themselves from other types of impressions; they have some kind of internal differentiating mark. Without this mark, we would have to rely on coherence between beliefs or some other such method. In such a case it would be inevitable that we would occasionally be wrong, and this would mean that we are not responsible for certain paths our lives go (because we were wrong with respect to some belief that influenced a choice, but being wrong about that belief was not our fault). This Stoics do not want to say this.

The only way to be completely responsible for our life is if we are able to avoid falsehood entirely.

Stoic Definitions of Cognitive Impressions

We get two versions of the definition of a καταληψις:
* SE’s and DL’s shorter version (two clauses)
* A more common, longer version (three clauses) The third clause is likely added from a debate as to whether CI’s have an internal differentiating feature.

The shorter version holds that an impression is cognitive if:
1) it comes from what is (apo huparchontos) and
2) it is imprinted and impressed in exact accordance with what is. (The impression must be clear and distinct.)

The requirement that it come from what is likely as the meaning of come from what is true (rather than come from a real object). To distinguish CI’s from other, noncognitive, impressions that are entirely true, the Stoics added the second clause. Then, because of debates with the Stoics, a longer definition was developed with a third clause:

3) it could not come about from what is not. “This, as we noted above, is taken to imply that an impression of this character could not be false…” (165).

Annas appears to take a different interpretation of (1) and (3), interpreting it to mean ‘from what is real’ and ‘from what is not real’, respectively. Indeed, if we look at SE’s account of CI’s, we see evidence to take ‘huparchein’ to mean ‘real’ rather than ‘true’. Consider 7.249: “The first of these is its coming about from a real thing; for many appearances strike us from what is not real, as in the case of crazy people, and these would not be apprehensive.” On the other hand, Cicero talks in terms of truth and falsity rather than existence. See 164-5 for Frede’s arguments in favor of the truth reading rather than existence reading.

The Criterion

How do CI’s constitute the criterion of truth?
* We can’t use them to certify the truth of any proposition.
* We can’t just consider some CI that corresponds to a belief that we want to test

Why? Because
* CI’s can only guarantee the content of their own propositional content
* If CI’s are only perceptual, they can only guarantee the truth of propositions that guarantee the truth of propositions that attribute a perceptual feature to an object. Also, there is no such thing as an impression that corresponds to a given proposition. One might have to go through a number of impressions to get to a CI. (You might have to go through a number of impressions of the man you’re walking toward until you get to a CI that the man is Socrates.) “Thus CI’s can’t be the criterion in the sense that we just have to look at our impressions to determine whether a proposition is true. It is, rather, by considering the proposition that we may get a clearer and clearer impression.
* Impressions aren’t pictures that can be examined introspectively but rather impressions are highly complex physical states; they’re imprints of objects in the world on our mind. So we cannot simply introspect as to whether the impression has the characteristics distinctive of CIs.

The distinctive mark that CIs have that marks them off from other impressions (and allows us to determine which impression is a CI and which isn’t) is the causal effect that it has on our mind. “The suggestion, then, is that the distinctive mark of cognitive impressions is a causal feature in that it makes the mind react in a distinctive way and that it is in this sense that the mind can discriminate between cognitive and noncognitive impressions” (168).

CI’s are called the criterion of truth because they can indirectly guarantee the truth of all other propositions known to be true. How? Because they give rise to general ideas (common notions; προληψεις) which in urn allow us to have further CI’s. For example, a CI of man leads to the common notion of man as a biped rational animal. This common notion is certified by (a) the CI’s which give rise to it and (b) the CI’s it gives rise to.

Looking at the Stoic account of common notions, it is clear that we don’t deliberately form common notions from CI’s. Our mind does this without our awareness of it. We just find ourselves with concepts that we didn’t have to start with.

Why does the mind assent to some impressions and not to others? There may be different answers, depending upon where one is in her cognitive development. Initially, it could be that CI’s cause the mind to accept them. Later, we may accept or reject impressions based upon the background of beliefs that we already have.

So it seems that the differentiating mark of CI’s is a causal, rather than a phenomenological, feature.

Cognition, Knowledge, and the Wise Man

The Stoics distinguish between knowledge, cognition, and mere belief.

Cognition: assent to a cognitive (in the wider sense) impression.
Opinion: assent to an impression that may or may not be cognitive
Knowledge: assent to cognitive impressions where the assent is of the sort of that firm such that one cannot be persuaded to withdraw the assent.

SO, one can’t have a false belief lest he be lead to accept the contrary of what he thought true via argument and dialectic.

All cases of cognition are cases of opinion or knowledge depending on whether they have the appropriate kind of assent. The average person will have nothing but beliefs because he cannot avoid any false belief and so his assent isn’t firm.

This leads the Stoics to say that no one has knowledge, while maintaining that it is still possible.

Skeptical Attack

The skeptics focused their attack on the doctrine of καταληψις. They took issue with the idea that a καταληψις (1) had a distinctive character such that there could not be an identical impression which was nonetheless false AND (2) by arguing that there are false impressions which have the characteristics of CI’s (vividness, etc) such that one could not distinguish it from a CI.

To (1) the Stoics respond that no two things are exactly alike and with training we can learn to distinguish the two objects. To (2), the Stoics can say that even in abnormal situations, we can distinguish between cognitive and noncognitive impressions and, further, it is a mark of deep abnormality if you cannot make such a distinction.

Whoever isn’t a sage is a fool (in much the same way that he who isn’t virtuous is completely vicious). There are two ‘cognitive achievement spaces’ – the cognitive achievement space that represents the certain, systematic, unshakeable space that is wisdom, and then the space that is everything else. We’re all in the ‘other else’ cognitive achievement space, as none of us are the σοφος.

Success and adequacy, then, is an all or nothing matter. This continues in the tradition of Socrates…since a person’s overall set of commitments is contradictory, no individual “plank” is safe.

But Socrates was notoriously rationalist (Plato? What is the Socrates that Meinwald refers to? She makes reference to the Republic here, as well as the Socratic elenchus more generally.) while the Stoics are empiricist. Wouldn’t this act as an important difference between Platonic epistemology and that of the Stoics. Might the empiricism of the Stoics lead us to hold that certain ‘planks’ (those from clear and distinct impressions) are safe?

Perhaps one way to respond to this is to note that επιστημη requires that one assent to the impression in a way that is unswayed by argument. If one has any false belief, then she can be swayed argument and thus has not assented in a way such that she can be said to have επιστημη.

We see this same position in ethics (the man drowning in 30 feet of water and the man drowning in 3 feet of water) and so it is natural to expect an analogous situation in Stoic epistemology.

Many scholars present a far more complicated account of Stoic epistemology, one that loses that starkness and clarity native to Stoicism.

Starkness and clarity are native to Stoicism? Consider Cicero in De Finibus (1.13) who says that he’s going to address Epicurean ethics first because that one is easiest to address (with the implication that Stoicism and Neo-Aristotelian accounts are much more complicated and more difficult to address).

An example of a complicated account of Stoic epistemology is provided by Annas (in “Stoic Epistemology”) who presents a series of forms of acceptances: assent, δοξα, καταληψις, επιστημη. She arrives at this ordering by looking at at Cicero’s recounting of Zeno’s fist analogy and SE’s account which says that καταληψις is between δοξα and επιστημη.

While neither account says that δοξα is above mere assent, Annas’ exposition does. But what sort of assent is weaker than Stoic δοξα?

Annas holds that δοξα and ignorance are distinct but finds that the Stoics blur the distinction between them.

This article will clarify how opinion should be characterized. Meinwald will argue for the restoration of a simple, clear-cut view with no form of assent below δοξα.

“Assent to an appearance generates a commitment which in the best case is knowledge and in all others is opinion (ignorance)” (218).

Look at Cicero Acad. 1.41: “What was grasped by sense perception Zeno called itself a sense perception, and if it had been so grasped that it could not be disputed by reason, he called it scientific knowledge, but if it were otherwise, he called it ignorance.”

Cicero Acad 1.42 = Zeno placed cognition between scientific knowledge and ignorance.
SE 7.151 = The Stoics say 3 things are linked together: scientific knowledge, opinion, and cognition between them.

One might think these are related in such a way that each characterizes a form of commitment corresponding to the contrast between the fool at the sage. Each has his own form of commitment. The sage has επιστημη and the fool has opinion/ignorance.
– this view is taken up in the middle books of the Republic.

I presume in book five when Plato says that knowledge is of things that are and belief is of things that are and are not. Or perhaps the allegory of the cave. Regardless, though, this doesn’t seem analogous. In the cave, the object of cognition is different – the lover of sights and sounds has as his object a particular thing of beauty while the philosopher has as his object of cognition the Form itself. Meinwald will go on to say that both the σοφος and the fool can have καταληψεις, presumably of the same thing, but because of their cognitive state (σοφος = stable, systematic, completely true beliefs; fool = not σοφος’s state) the καταληψις for the one will be knowledge and for the other mere opinion. This doesn’t seem analogous to Plato.

Just as any individual is either a sage or a fool, so too any mental disposition is an instance of επιστημη or δοξα/αγνοια

“Each character would have his proprietary form of commitment. And for the most clear-cut view each of these forms of commitment would be made up of eponymous parts, i.e. knowledge as a state would consist of bits of knowledge, and opinion=ignorance as a state would be made up of bits of opinion=ignorance” (219).

But the work of most scholars goes against this. Why? Because it is generally thought that an individual opinion is assent to an incognitive impression. Once you accept this account, you give up the simple picture in one of two ways:
1) you disturb the relation between the state of opinion and individual opinions
2) you take opinion and ignorance to be different states.

Gorler adopts view (1). He says that someone has the disposition of opinion if some individual commitments are opinions (ie assents to non-kataleptic impressions). This means that some of the commitments that make up the disposition aren’t themselves opinions.

But this leads us to worry about the Stoics not making it clear if they’re talking of the state of opinion or particular opinion. They’re vague about this in places, and if they actually held the view as advocated by Gorler, you would expect them to be much more precise in their use of the words.

The more mainstream option is (2) where ignorance and opinion come apart. Both Annas and Long & Sedley make δοξα cover assents to the incognitive while ignorance extends to all commitments that fall short of knowledge. So opinion consists of all and only particular opinions (while ignorance is the state).

This gives up on the binary set-up that seemed characteristically Stoic. Further, it doesn’t take into account (or, perhaps, does not adequately take into account) that the Stoics sometimes do seem to identify opinion with ignorance.

What we should do is review our evidence for taking ‘assent to the incognitive’ as the mark of opinion. (I’ll call it the AI argument)

* * * *

A central bit of evidence in favor of the AI argument is a bit of Plutarch. Looking at this bit of Plutarch (On Stoic Self-Contradictions; LS41E) we see that if one opines, she yields to impressions which are incognitive.

This passage gives us the conditional ‘if one assents to the incognitive, then she opines’ but it does NOT give us the conditional ‘if one opines, then she assents to the incognitive’.

Another passage LS point to is at 41G which commits us to the position that since the sage doesn’t opine, he doesn’t assent to anything incognitive. This commits us to the claim that if the sage were to assent to the incognitive, then he would opine. It does not commit us to AI.

CM proceeds to address other passages – Cicero Acad 2.59 and 2.60; and LD 41D2 – that LS point to to support AI. CM argues that they fail to offer adequate evidence for the claim.

We may take Acad 1.41-2 to lend support for AI. But we need not. The argument that Arcelaus presents is as follows (CM pg 226):
(1) cognition occurs either in a sage or in an inferior man
(2) cognition in a sage is knowledge
(3) cognition in an inferior man is opinion
(4) there’s no other possibility
(5) so there’s nothing between knowledge and opinion to be the criterion

If we look at SE 7: 153-4, we see evidence for the simple reading. As presented, if we take the complex interpretation (the one advocated for by LS and Annas), then Arcesilaus says something false (namely, if a cognitive impression occurs in an inferior man, it is opinion) since that view says that opinion is defined as assent to the incognitive (and Arc. says that cognition in an inferior man is opinion).

The simple reading can handle this argument, but it must account for P5: there is nothing between knowledge and ignorance to be the criterion.

Whether (5) follows from (1) – (4) depends on the interpretation of ‘between’ that we use. We can’t mean it like paving stones, where one paving stone is between two others, because then knowledge couldn’t be opinion and SE goes on (at 7.151) to describe it thusly. Also, cognition is said to be common to both the wise and foolish.

We should think of ‘between’ as a sort of yoke connecting two things (like two oxen yoked together). This lends support to the simple reading since cognition figures in to both knowledge and opinion.

If we take ‘between’ in this way and go back to P5 we see that A’s argument is neither embarrassing failure or a knockout (because it doesn’t show that cognition can’t be between knowledge and opinion) and so the Stoics still have room to respond to Arc.’s worries. This preserves the dialectical requirement as put forward by Annas in “Stoic Epistemology”.

Stoic epistemology addresses both ancient and contemporary concerns (concerns with both understanding and the nature of knowledge). Like the Epicureans, they’re empiricist. Knowledge is based on appearances.

Appearances are how the world appears to us. They make an imprint on us. This imprinting is a mental and physical act. Further, it is structured in a way that the mind can articulate.

In any perception, there will be
- an appearance, and
- some kind of acceptance of the propositional content of the appearance

The weakest sort of acceptance is assent (συνκαταθεσις)
The next strongest sort of acceptance is belief (δοξα)

Belief is assent to what is not apprehended (either what is false or rash assent to what is true).

“The latter is presumably assent to what is true, but as far as the person goes, might have been false, since his assent was rash and did not spring drom the firm and systematic grasp of the subject matter characteristic the person of knowledge.” (186)

But the Stoics want to say that anyone can have apprehensions, even those without systematic grasps charactistic of the person with knowledge. If the requirement for not having rash assent is that it springs from a firm and systematic grasp of anything, then very few people will not always engage in rash assent. Why can’t we just say that this is rash assent because the appearance is not apprehensive, and leave it at that?

Knowledge excludes belief and belief is a weak assent. We can understand this in two ways: first, the assent itself is weak (contrast with apprehensions which drag you to assent) and second, the belief is an assent made from weakness (the person lacks coherance, stability, and system).

The next stage is apprehension (καταληψις)

Apprehension is the stage where you cannot be wrong. It is assent to an apprehensive appearance (φαντασια καταληπτικη)

The final stage is knowledge (επιστημη) in which particular facts are grasped systematically, coherently, etc.
– Only the σοφος is capable of this
– It is unchangeable by argument
– It is a building up of beliefs
Knowledge is the culmination of the process which begins with one’s reactions to appearances.

One comes to form beliefs via appearances. One builds up a set of beliefs from those appearances, ensuring that they are systematic, consistent, coherant, stable, and so forth. It is only through this that one comes to have knowledge, and it is in this way that knowledge is a culmination of the process that begins with appearances.

Apprehension is the stage in which you cannot be wrong, and so 1) can be thought of as a sort of knowledge and 2) is the criterion of truth.

Because apprehensive appearances are so crucial to Stoic epistemology, we need an account of what they are. We get that in Sextus Empiricus 7: 241-58. Here we discover that an apprehensive appearance (AA) is:
1) from an in accordance with a real object
2) stamped and sealed
3) could not have come from an unreal object
4) older stoics say that it is the criterion only if the appearance has no obstacle

While S.E. expresses the definition of AA in terms of the relation between an appearance to a real or unreal object, Cicero says that what was important was not the object’s existence but rather the way that it is represented.

S.E. says that an AA represents the object in every detail while Cicero says that Zeno says that it represents objects only in part. (This will be important later).

AA’s are normal. Anyone can have them so long as the conditions are normal. The five conditions of normalicy are:
1) conditions of the sense organs
2) condition of the object
3) the placing of the object
4) the way the object is sensed
5) the agent’s state of mind

AA’s are representational objects – they are caused in a way that represents the object to the person’s mind.

The Stoic theory places a represntation object (the appearance) between the agent and the world:
agent appearance object in world
There’s a concern about the representational object bridging the gap between the agent and the world. Why should we be confident that the appearance accurately represents the way things are.

This appears to leave the Stoics open to a worry: why are we entitled to being confident that things are really as they are represented as being?

Note the difference here between the Epicureans and Stoics. The Epicureans also face this criticism and respond that there is no representation between the object and the world. The effluences hitting the eye just ARE the objects in the world that directly cause the perception in the agent. The Epicureans are able to avoid this issue, then. The Stoics aren’t.

We have two Academic arguments preserved in Cicero and S.E.:

Argument #1
There can be appearances with all the distinguishing marks of an apprehension (being striking and evident) but which nonetheless are false. (Consider the visions of a madman)

The Stoics just deny this and claim that there IS a difference 1) either phenomenologically or 2) in the state the experience is had in.

The Skeptics reassert that there is no difference.

Argument #2
There are some objects that we cannot tell apart (two eggs, for example). So someone can have an AA of one egg and thereby has an indistinguishable appearance of a different object (the other egg). This leads us to conclude that an AA cannot possess a mark which distinguishes it from something which is like it. (See S.E. 7:408-11)

The Stoics respond to this in two ways
* No two things are exactly alike and so some difference will be reflected (the Stoics have an account of the identity of indescernables, thus this isn’t just an ad hoc response)
* We could, if we tried hard enough, distinguish between the two seemingly indescernable things (two eggs, for example). Just because we can’t usually distinguish between two things doesn’t mean they’re actually indistinguishable.

Because of the two Academic arguments (above), we see three changes in the Stoic account.
FIRST, the final clause of the definition was added
SECOND, Zeno initially claimed an AA didn’t represent the object in every detail
THIRD, the later Stoics added the clause ‘being no obstacle’

FIRST CHANGE: THE FINAL CLAUSE
The Stoics added that an AA was such that it couldn’t come from an unreal object. This appears to be in response to argument 1 (above).

But the Stoic response to argument 1 demands some response. What is the difference between the objects (because they claim that there will always be a difference)?
– is it a phenomenological difference?
This is a weak response that appears to miss the point of the criticism.
– is it that the AA is distinguished by its causal history (ie that it is produced in a normal state)?
This response would settle the matter…but the debate continued. The skeptics future responses would be irrelevant if this is the proper way to interpret the Stoics’ response.

Perhaps Zeno’s position was indeterminate. Perhaps he didn’t sharply distinguish between two key ideas:

IDEA A: Apprehension requires that a person is in the right relation to the object known

IDEA B: Apprehension requires a person is in the right relation to the object known and this fact is available to her.

The skeptics are pressing on idea B.

We see that this debate between the Stoics and skeptics revolve around the question of what is sufficient for knowledge.
- if idea A is, then the Academics seem to miss the point with their criticisms
- if idea B is, then the Stoics seem to miss something that the Academics are pressing them on.

We see, then, that there is a disagreement about what knowledge requires.

SECOND CHANGE: PARTIAL OR TOTAL REPRESENTATION
S.E. says that an AA represents every feature; Zeno says that is represents only some features.

Argument #2 seems to motivate this change. If an AA represents the object in every detail, and no two objects are exactly similar, then an AA will have a distinguishing mark.

But the debate DID continue…so the answer cannot be that simple.

The response relies on the identity of indescernables and the claim that the distinguishing mark can be discovered, at least by experts.

The Stoics don’t say that no one save the σοφος can discern the distinguishing mark. It is perplexing why they don’t. Why do they refer to Delian poultry farmers and the like rather than say that the two objects are discernable but only the σοφος can disern them?

This hesitancy on the part of the Stoics uncovers a serious philosophical question: “Is it sufficient for knowledge that I should merely be in the right relation to the object of knowledge [idea a, above], or is it also required that this relation should be something available to me [idea b, above]?” (199)

The Stoics begin with idea b (via Zeno’s demand). But the Skeptics respond with argument #2. This leads the Stoics to strengthen the condition for an AA, although they must tacitly weaken the conditions for an AA from idea b to idea a. The Academics seems to continue to press the issue, though, which indicates that the Stoics were indeterminate between (a) and (b).

Zeno’s original demand: an AA should enable the person to distinguish its object. The Stoics (μεν) strengthen the condition for an AA by saying that it represents every feature of the appearance, but also (δε) weaken the condition by saying that it can be an AA only if you are in the right relation to it (you yourself need not be able to discern the difference). This appears to remove Zeno’s original demand (right??) and shift from (b) to (a).

THIRD CHANGE: REMOVAL OF AN OBSTACLE
The Stoics must have been unhappy to have been forced from idea (b) to idea (a) as it means that someone can have a criterion of truth but be unable to use it. Because of this, we see later Stoics adding that an AA is a criterion of truth only if there is no obstacle. This suggests unhappiness with the idea that one has a criterion of truth but is unable to use it.

We should take the reference to exterinal situations to refer to a normal appearance where a person’s beliefs form an obstacle to accepting the AA as such. BUT, we have to rely on far fetched stories for this (ie Menelaus and Helen in Euripides’ play). This is problematic because
* we must bring in powerful and irresponsible gods (or something similar)
* this doesn’t help in cases where someone is in an abnormal experience and doesn’t realize it.
* Apprehension seems to love its position as a kind of knowledge. We don’t seem able to distinguish it from επιστημη any more. Apprehension was originally supposed to be available to all, but now it looks to demand a consistent and coherant system of beliefs. (So it is like επιστημη and it will not be available to all.)

Conclusion:
The Stoics begin with a common sensical theory but are indetermined as to whetehr apprehension requires idea a or idea b. (b) is more intuitive and the skeptics press them on this. This forces the Stoics to retreat to (a) but in doing so makes their account less common sensical and it renders AA’s an unusable criterion of truth.

With re. to (2): kritèrion is used to refer to a courthouse or law court

The philosophical term ‘kritèrion’ is related to the Attic kritès which means ‘evaluator’ or ‘arbitor’.

The term ‘kritèrion’ doesn’t specify the type of instrument or method used. Indeed, we find diverse things being called kritèria tès alètheias.

Because many different philosophers considered a diverse range of things to be kritèria tès alètheias, we cannot use a translation that assigns the criterion a specific function. We cannot say, for example, that it is a ‘distinguishing mark’ or ‘standard of truth’ as those denote specific functions which may not apply to all things called kritèria tès alètheias.
– we may be able to be more concrete in our translation of ‘kritèrion’ for particular times or particular schools.

Uses Before Epicurus

Plato at Rep 582a6: “By what should a thing be judged if it is to be judged correctly?” and at Tht 178b6: “Man is the measure of all things…for he has in himself the criterion for these things.” Aristotle uses it once at Met. K6 1063a3.

In Plato, criteria are 1) qualities or capacities (Rep.) or 2) faculties (Tht.) which provide an ability to evaluate the truth or falsity of 1) particular assertions (Rep.) or 2) opinions (Tht.).

In the Tht. passage and the Met. passage, ‘criterion’ is used in connection with the Protagorean thesis. This might be evidence that Protagoras initially used the term.

Epicurus

Epicurus did not use the term ‘kritèrion’ only as a term for a cognitive faculty because while aisthèsis might refer to a faculty, prolèpsis cannot be construed as a faculty. So E. used the term ‘kritèrion’ in at least two different uses. Striker is going to focus on this second use.

Kritèrion and Kanwn

Kanwn refers to a straight stick or rod to be used as a measuring tool. It is a tool for testing the straightness or crookedness of something and so the tool must itself be straight.

Using the idea of a kanwn, we should expect two things of a means of judgment:
1) it should itself be true
2) it should be able to serve as a means of testing truth

The Truth of Criteria

The truth of a criterion cannot be determined or tested in the same way as the truth or falsity of what is tested by it. The truth of a criterion must be assumed to be self-evident and not in need of verification. (Cicero mentions is (CITE?) and it is in KD XXIV)

There are two criteria: 1) aisthèseis and 2) prolèpseis

Aisthèsis
“Since Epicurus had claimed in another passage that there could be no criterion superior to perception, he was compelled to assert that all perceptions were true – ie self evidently true – or else to renounce the possibility of distinguishing truth from falsity” (34). If we accept this, then the truth of the senses cannot be taken as a fundamental postulate of Epicureanism.

It appears that the argument goes as follows: Epicurus was compelled to posit the absolute truth of perceptions because he claimed that perception was a foundational criterion (such that there could be no superior criterion). From this we see that the truth of the senses cannot be a fundamental postulate of Epicureanism.

I worry about this argument. Can’t we also say that Epicurus was motivated by his commitments regarding atomism and perception? While he does say, several times, that the senses must be true if we are going to avoid an infinite regress in confirmation or a collapse of truth and falsity he also provides a physical theory as to why the senses must be true. Why should we think that the epistemic motivations (regress, collapse of t & f) should take theoretical precedence in explaining Epicurus’ claims regarding the truth of the senses?

When E. talks of perceptions being true and false, we should understand him to mean ‘perceptions that (such and such is the case)’. “A perception is true when the proposition which expresses it is true” (35).

I’m not convinced. Why should we think that Epicurus MUST have been referring to the truth and falsity of propositions here? Why can’t we say that the Ancients may have had a different sort of notion of truth than we do, one that is not restricted to propositions? Plato talks of the Forms as being true but we can’t (or, at least, shouldn’t) say that the Forms are propositions. I’m simply not convinced that when Epicurus says that perceptions are always true that he cannot have simply meant the perceptions themselves are always true.

But what does it mean to say that a proposition expresses a perception? Which propositions can be taken as expressions of a perception?

Three different types of propositions have been taken as expressions of perceptions:
1) propositions concerning specific sense-objects (this is hot)
2) propositions concerning observable states of affairs
3) empirical generalizations
Epicurus appears to have used all three.

Epicurus clearly appeals to propositions of type (1). But these can’t act as a sufficient basis for knowledge of the perceptible world.

So Epicurus introduces the notion of ‘phantastikè epibolè tès dianoias’ to make use of propositions of type (2). ‘phantastikè epibolè tès dianoias’ involves focusing one’s mind on an object or state of affairs.

It’s hard to see how Epicurus justified the inclusion of propositions of type (3), as any arguments in support of the claim that all perceptions are true refer to individual perceptions.

Prolèpsis
There is an argument for the self-evidence of prolèpsis at Letter to Heroditus 37-8

The argument says that prolèpsis are requites
1) so that we may have a reference point against which to judge matters of opinion, inquiry and puzzlement
2) without prolèpsis, demonstration will go on endlessly
3) so these criteria are ncessary prerequisites of a demonstration

These prolèpseis must necessarily be grasped without further explanation if they are to serve as criteria.

Further, this infinite regress argument rests on TWO key facts:
1) we cannot rely on prolèpsis as a criteria
2) these criteria must be proved
It’s only if we accept both of these that we get an infinite regress.

This tells us something about prolèpseis: they are describable as demonstrated or are indemonstrable.
– so they are something which can be true or false and so cannot simply be mental images.

Prolèpseis are referred to 1) as seen (and so it is natural to conceive of them as images) and also 2) demonstrated or indemonstrable (and thus understood as propositions)

So we have an account of the notion of a prolèpsis: it is a self evident truth which must have been grasped at the outset of an inquiry.

Methods of Testing Beliefs

Criteria are used to test non-evident truths or falsehoods

A doxa is true if
1) it is confirmed (epimarturèsis) by enargeia
2) it is not refuted (ouk antimarturèsis) by enargeia
A doxa is false if
1) it is not confirmed (ouk epimarturèsis) by enargeia. (This occurs when the perception not-p occurs, not simply when the perception p does not occur)
2) it is refuted (antimarturèsis) by enargeia

(1)’s deal with beliefs that are prosmenon
(2)’s deal with beliefs that are adèlon

An adelon is false if it is refuted by an enargeia (the negation of a perceptional proposition follows from the adelon)

An adelon is true if it follows from some self-evident perceptual proposition.
– It can’t be sufficient to be compatible with the phenomena because that will allow opposite claims to both be true.

“An opinion concerning observable states of affairs which are neither confirmed nor not confirmed by perception counts as possibly true” (48).
– one opinion must be true, but we can’t determine which one. So when E. talks of something occuring pleonaxws, he isn’t saying all these ways ARE true but just that they are all POSSIBLY true.

A prolèpsis acts as a criterion in two ways:
1) Cases of the form ‘x is y’, the prolèpsis of x can be used to test whether the predicate (Y) follows from the prolèpsis of X.
2) In judging an opinion of the form ‘y is x’ we can test whether y has the qualities which an x (according to its prolèpsis) must have.

The Stoics and Skeptics

The stoics named several different things criteria. The official view says that the criterion of truth is a katalèptikè phantasia. A cataleptic impression is one that “comes from what is, is imprinted and impressed in exact accordance with what is, and is such that an impression of this kind could not come about from what is not” (51). They bear a ‘characteristic mark of truth’ by which they can be distinguished with certainty from false impressions.

What is meant in calling something like this a criterion of truth?

A katalèptikè phantasia is a criterion for the very state of affairs from which it arose.

The criterion isn’t an instrument to judge the truth or falsity of opinions (like E’s) but as a means to establish what is the case within the scope of perception.

The katalèptikè phantasia gives the conditions which must be met by an impression which leads to knowledge. If this impression can be characterized independently of its relation to the state of affairs which occasions it (by the ‘mark of truth’) then it could be taken as a sign of the existence of the SoA in question.

The same holds (basically) for the Skeptics (esp. Carneades), except:
– The Stoics hold that the criterion was a guarantee of the SoA
– Carneades denied that there could be such a guarantee

Chrysippus also said that prolèpsis and aisthèsis were criteria. In another book he named katalèptikè phantasia as a criteria. How to deal with this?

We can try to accomodate them by saying that prolèpsis and aisthèsis are elements which play a role in the formation of a katalèptikè phantasia. In this way, the use of prolèpsis as a criterion is the same as Epicurus’.

I’ve also switched from a blogger account to a wordpress account. I like the options available on wordpress that simply aren’t available (at least easily) on blogger.

Also, if anyone who reads this (and god knows that’s likely no one) knows how to post things in Greek fonts, I’d gratefully receive tips. I’m starting to hate doing translaterations.

I.
The PA does two things:
1) provides a theory of the structure of science; the conditions for a proposition to belong to a body of systematic knowledge;
2) gives an account of the cognitive state of an individual who has mastered a body of systematic knowledge.

Aristotle first gives an account of the cognitive state of epistasthai (knowing the cause and necessity) and then draws the second conclusion about the objects of episteme.

“Because episteme involves grasping the demonstration of necessary conclusions, it is grounded epistemologically on the premises of that demonstration” (99).

Aristotle is concerned with one’s cognitive state with regard to particular propositions (rather than who branches) because he distinguishes between unqualified episteme with respect to a theorem of science from various qualified versions of episteme in relation to the same theorem. (100)

What should we make of the distinction between qualified and unqualified episteme?
* qualified is still episteme, just not the favored kind.
**How should we make sense of this? Demonstration from archai is only one form of proper justification; something can have justification without deductive demonstration. (See A2, A5, and A13 for places where Aristotle recognizes this distinction.)

“In other words, Aristotle both knows and emphasizes that his requirement that demonstration proceed from first principles is not a requirement of justification but of scientific explanation.” (101)

(NOTE: missing pp 102&3)

“In the PA, episteme is coordinate with epistasthai and denotes either a cognitive state or the body of knowledge he has mastered” (105).

In A2, Aristotle says that epistasthai is ordinarily conceived that: “x epistatai y iff (a) x gignoskei the explanation of y is and (b) x gignoskei that y cannot be otherwise than it is.”

This fits our conception of UNDERSTANDING much better than it fits the concept of knowledge. This isn’t to say that we can’t use the term ‘knowledge’, but we have in mind the sense of the word when we say ‘he has knowledge of mononucleosis’ rather than ‘he has knowledge of where is car keys are’.

“Aristotle is analyzing a cognitive state which is achieved by knowing explanations, and whether he is currently calling it episteasthai or gignoskein the corresponding term for that philosohical state in philosophical English is ‘understand’” (107).

II.
Episteme is of what cannot be otherwise.

“If Aristotle is making a claim about understanding, hiw point will be that understanding depends on explanation and what gets explained in the sciences…which general regularities and connections: lawlike regularities in the modern jargon, necessary connections in Aristotle’s” (109).

Doesn’t Aristotle also talk of things that happen “for the most part” and wouldn’t this be the law-like regularities?

Burnyeat references the EN here. The quote from the EN is as follows: “Clearly then practical wisdom is a virtue and not a skill. And sicne there are two parts of the soul that possess reason, it will be in virtue of one of them, namely, that which forms beliefs, both beliefs and practical wisdom being concerned with what cannot be otherwise. (EN 6.5, 1140b).”

One can apply the explanation of a recurring type of phenomenon to a particular instance of it; but this results in accidental episteme, not episteme haplos.

Scientific explanation is at first an explanation of laws rather than an explanation of particular events.

When Aristotle says that knowledge is of what cannot be otherwise, he’s commenting that understanding is constituted by knowing the explanation of necessary conditions in nature.

The considerations make more sense as considerations about explanation than considerations about knowledge or certainty.
1) because epistasthai involves explanation that Aristotle insists on the characteristics of the principles (true, primitive, etc)
2) necessary premises is a requirement of explanatoriness.

Burnyeat points to A6 for this. I don’t see it here, though. What A says is that if he apodeiktihe episteme comes about, you need necessary principles. Does Burnyeat want to say that apodeixis = explanation?

Aristotle’s most substantial claim (at a.6, 74b26ff) is that “to explain the holding of a conclusion that is necessary one must demonstrate it through a necessary middle term” (110).

To understand a theorem you must understand
1) that it is necessary
2) WHY it is necessary (it is necessary because it id demonstrable from prior principles which are necessary, and those prior principles are necessary because they are per se predications expressing a definitional connection)

It sounds weird to say one understands THAT something unless the sort of understanding you’re talking about is semantic understanding.

In B.6, although the details aren’t clear, “it seems fair to say that he is trying to give substance to the idea that the fundamental predications of a science ought to be self-explanatory. (ie they’re not merely immediate (no explanation via a middle term) but actually explain themselves).

Aristotle is looking for substantive knowledge of a thing’s essence in a scientific definition.

The man who achieves unqualified episteme is a man for whom every “why?” question in a given domain has been given an appropriate answer. His grasp is:
1) systematic and
2) synoptic
In that everything in the domain is explained in light of first principles which explain themselves.

Aristotle has a vision of complete understanding which supports his claim that one can only have episteme of what is universal, necessary, and everlasting.

Aristotle is not saying that we can’t KNOW that accidental states of affairs obtain; it’s just that accidental knowledge falls outside the reach of systematic explanation and understanding.

There is no episteme through perception of particular things because explanation requires generality and this is beyond the scope of explanation. Perception DOES yield knowledge, though. (gnosis, not episteme)

If we understand episteme as justified true belief, then Aristotle presents a very skeptical view. If we take it as understanding, then the restrictions he places upon it are intelligible.

III.

Aristotle doesn’t appeal to concepts like evidence, certainty, and justification.

“This is reason, of course, for disavowing the once prevalent idea that the PA advocates demonstration as a method of scientific discovery”. Rather, demonstration is a method of teaching facts already won. It is not how you discover facts but how you should present and impart them.

But it is bad pedagogy to lead a pupil straight to archai and then launch into a remorseless chain of syllogisms.

How do we make sense of the PA as good pedagogy, then?

Consider again the distinction between knowledge and understanding.
* When you teach by imparting KNOWLEDGE, you will include evidence and justification in your teaching.
* When you teach via imparting UNDERSTANDING of knowledge that pupils already have or understanding of a science with which they are already acquainted in an unsystematic way, one does this by unifying things, putting them into a systematic framework.

Think of this education (put forward in the PA) as an advanced graduate level course, not an introductory course.

IV.

Does this account of Aristotle’s pedegogical philosophy improve the claims of demonstration?
*it depends on whether we agree on certain philosophical questions regarding understanding.

The key to understanding (for Aristotle) is demonstration and the sort of demonstration that Aristotle has in mind is demonstrations of the form Barbara (AAA. AaB, BaC, therefore AaC)

Even if we grant that understanding can be gained from relating and organzing bits of knowledge, it doesn’t follow that understanding is to be sought from putting knowledge into ARISTOTLE’S mould (ie syllogisms and demonstrations).

The extent to which we doubt understanding is the fruit of axiomitization is the extent to which we doubt that demonstration is the method in which to impart understanding.

“A teacher can sensibly aspire to conduct Aristotelian demonstrations if it is right to claim that, where we can achieve full axiomitization will provide us with a completed structure of explanation which should be the idea fulfillment of a common conception of understanding” (126).

“For Aristotle, an axiomatic system isn’t just a preferred ordering of humanly constructed knowledge, but a mapping of the structure of the real” (126).

V.

In A2, A. says that it is a requirement of episteme that one know and be convinced of the first principles. Can’t this be so I can know and be convinced of the demonstrated conclusions?

Explanation, being prior to what it explains, is more knowable and familiar in the order of nature, and is thereby more believable and convincing.

The relationship between the knowable (familiar) and the convincing is important.

Aristotle makes a distinction between what is more knowable in the order of nature nad what is more knowable to us.

Demonstration produces gnonai…but this gignoskein is knowledge as a grasp of what is knowable by nature.

So we can constrast knowledge with and knowledge without full understanding.

You can have knowledge and even have put things together in an orderly way and STILL not have mastered them because you lack intellectual practice and familiarity. (in EN VI and VII, ee B. pp 129-30).

Aristotle didn’t see a problem transforming inductive belief into knowledge (gnosis) but he thought that this isn’t yet understanding. To get understanding we need more familiarity, more dialectical practice, intellectual habituation.

The archai are knowable in themselves, etc; when one has propositions we knows on inductive grounds which are convincing and knowable in themselves, one need just become fully familiar with them and convinced. This conviction and understanding is nous. (132)

The central claim of this chapter is that P is demonstrated through necessary principles.

1. If demonstrative understanding proceeds from necessary principles; and
2) if whatever holds of an object in itself is necessary, then
3) demonstrative deduction will proceed from certain items of this sort (ie necessary items)…because everything holds either incidentally or in this way (presumably through demonstrative deduction)

We must argue like this or posit as a principle that demonstration is necessary. “for from truths you can deduce without demonstrating, but from necessities you cannot deduce without demonstrating.”

There is evidence for this idea that demonstration proceeds from necessity because we say that the premises aren’t necessary if we think it possible for them to be otherwise.

A further argument in favor of the idea that demonstration must proceed from necessities is as follows:
1) if in a case of demonstration, someone who doesn’t possess an account of the reason why doesn’t have understanding; and
2) if it might be A holds of C by necessity but B (the middle term through which this is demonstrated) do not hold from necessity, then
3) he doesn’t know the reason why.

If I ‘demonstrate’ the necessary proposition AaC from non-necessary premises, then I don’t know why AaC and thus do not understand AaC

If premises are contingent, then AaC can’t hold because of them.

This seems to indicate that the way to come to understand the reason why comes via demonstration with necessary truths. Hopefully later he’ll give an account of how demonstrations act as a proper way to come to know the reason why. Or why the results of demonstration are considered accounts of the reason why.

It’s also interesting to note the idea that understanding does not come about without having an account of the reason why.

If someone does not know something now
1) although he has an account,
2) the account is preserved (ie the person is alive)
3) the object is preserved (the truth; p is true at t1 and at t2)
4) he has not forgotten
Then he didn’t know it earlier.

A presupposition apparent in (3) is that a proposition (P) might change its truth value between t1 and t2.

Would this be for things that happen for the most part? Necessary truths can’t change their truth value, surely.

But the middle term might perish (be false) if it isn’t necessary.

But because the person is preserved, and the object is preserved, he will possess an account, and yet he doesn’t have knowledge, and therefore he didn’t have knowledge earlier.
* even if this hasn’t occurred, it can (because the middle term might perish) and so one can’t have knowledge in these conditions.

Even if a conclusion holds of necessity, the middle term need not because you can deduce a necessary truth from non-necessary truths (in the same way you can deduce a truth from non-truths).
*BUT when the middle term holds from necessity, the conclusion also holds from necessity (like a conclusion from true propositions is always true).

Since understanding from demonstration must hold from necessity, your demonstration must proceed from a middle term which is necessary. Otherwise you will understand neither the reason why nor that it is necessary for it to be the case.

Of incidentals which do not hold of things in themselves, there isn’t demonstrative understanding because you can’t prove the conclusion from necessity (because it is possible for the incidental to be otherwise).

One might wonder why one would even ask about incidentals if it isn’t necessary for the conclusion to hold.

For any argument of the form P so necessarily Q will be good since NO argument of that form will be good, where Q is not necessary.

We should ask these questions NOT because the conclusion will be necessary because of the points proposed in the question but rather because it is necessary for anyone who accepts the proposals to state the conclusion (truly).

Since whatever holds of something in itself and as such holds of it by necessity, it is clear that demonstrations must proceed from such items (in itself & by necessity). This is true because what is incidental doesn’t hold necessarily and so you don’t know necessarily that the conclusion holds (even if it is always the case)…because you will not understand in itself
1) that it holds in itself, nor
2) why it holds

We often make mistakes. What we try to approve does not hold primitively and universally, even though we think we are proving it universally and primitively.

We make mistakes:
1) when there is nothing higher we can take apart from a particular case;
2) when there is something higher, but it is nameless and covers objects of different forms
3) when the proof applies to something that is a partial whole

Example of 1: If the only triangles we were aware of were isosceles, we might believe 2R held universally of the isosceles because we have no concept of triangle.

Example of 2: Aristotle is referring to a universal mathematics – an allusion to Eudoxian generalization.

Example of 3: If we observe that all perpendiculars are parallel and we wrongly inter that this is what we demonstrate.

When do you not know universally but do know simpliciter?
* you would know simpliciter if it was the same thing to be a triangle and be equilateral. But if it is different and if something holds of them as a triangle, then you don’t know it.

To what does the demonstration apply universally?
* to the first item after the removal of which it doesn’t hold.
— for example, 2R holds of a bronze triangle. When ‘bronze’ is removed, it is still a triangle. But when ‘figure’ is removed, it is no longer a triangle.

But ‘figure’ and ‘limit’ aren’t the first (item after the removal of which…), rather triangle is the first.

‘first’ means first terms after whose abstraction R fails to hold. The thing that holds first holds primitively; so the abstraction colds primitively of ‘triangle’.

I found this to be a nearly incomprehensible chapter. The notes below barely make sense, Barnes’ commentary, while helpful, was also baffling (although I doubt Barnes is at fault here). This is something I need to go back to at a future date when I might have a better idea of what is going on.

PA, Alpha, Chapter 4

Because it is impossible for understanding simpliciter to be otherwise, demonstrative understanding will be necessary.

A demonstration is a deduction which proceeds from necessities. SO we must see what kind of items demonstrations proceed from..

We must define three things:
1) ‘of every case’
2) ‘in itself’
3) ‘universally’

Something holds of every case if it does not hold in some cases and not in others, nor at some times and not at others.
* for example, if ‘animal’ holds of every man then if this is true to call this a man, it is true to call him an animal, and if he is now he also was previously.

Something holds in itself if it holds in it what it is.
* for example, the line of triangles or points of lines. (The essence of a triangle comes from the lines.)
If what it holds in itself inheres in the account which shows what it is, there is an account which specifies the specific thing it is.
* for example, ‘straight’ and ‘curves’ both hold of lines; ‘even’ and ‘odd’ both hold of numbers.

Aristotle gives four ways A can hold of B ‘in itself’. The first two:
1) A holds of B in itself = df ‘A holds of B and A inheres in the definition of B’
ex: animal holds of man in itself
2) A holds of B in itself = df ‘A holds of B and B inheres in the definition of A’
ex: mortal holds of animal (good example?)

Barnes then talks of I-predication. A proposition is an I-predication if
a) it is of the form ‘Every B is an A’ and
b) it is true in virtue of the fact that A holds of B in itself.
The proposition is an I1 proposition if ‘in itself’ is taken in sense 1 (above) and a proposition is an I2 proposition if ‘in itself’ is taken in sense 2 (above).

Certain items are not said of some other underlying subject. Things that aren’t said of an underlying subject are called things in themselves.

A. distinguishes things that exist in themselves (independently) with things that exist incidentally.

A distinction between natural and unnatural predications. If ‘x is y’ is a natural predication, then
1) ‘x is y’ doesn’t entail something else is Y and happens to be X
2) X is not ontologically dependent on anything else, and
3) x is an independently identifiable subject of change

Another way to describe that something holds ‘in itself’ is to say that what holds of something because of itself holds of it in itself.
* for example, if something died while being sacrificed, it died in the sacrifice since it died because of being sacrificed and it was not incidental that it died while being sacrificed.

This involves a connection between events

When something is understandable simpliciter and is said to hold of things in themselves either by
1) inhering in what is predicated or
2) being inhered in
then it holds of them because of themselves and from necessity. This is because it is not possible for them not to hold
1) either simpliciter or
2) as regards the opposites.

A. is referring to I1 and I2 predications. When A. says that “A holds of B simpliciter” he is talking of I2 predication, and when he refers to “as regards the opposites” he refers to I2 predications.

‘Universal’ is defined as what holds of every case in itself and as such.

Something holds universally when it is proves of an arbitrary and primitive case.

The definition of ‘universal ‘has three components:
1) ‘of every case’
2) ‘in itself’, and
3) ‘as such’ (qua)

A holds B as such iff there is no term C which explains why A holds of B; iff AaB is immediate.

Being an I-predication and holding as such are logically equivalent.

reread Barnes’ summary of the chapter on pp 120-2. Are there any articles that might make this a bit more clear?