Preface

In Spring Quarter of my third year, I took a course on Epistemology under Duncan Pritchard.  For the final essay, I responded to Pritchard's views of 'knowledge' and 'understanding'.  As is usual for my essays, I wrote this one in the 36 hours before the deadline and eventually turned it in a bit late.  (Don't worry, I still did well.)  What is posted here is a revised version of that essay.

The prompt that I chose asked me to show the relationship between knowledge and understanding.  Much of the essay is devoted to solving an apparent discrepancy in Pritchard's account of understanding.

Finally, as is the case with most of the essays I write for classes, I rarely believe most of what I write.  In this case, I'm not a fan of analytic philosophy when it attempts to define words because I'm convinced that there is no complete definition for these evolving social constructs.  Even if we were to find one, the word's usage would likely change later.  It doesn't make sense to me to argue about which concept should be raised on the pedestal of being the definition - the concepts aren't any more or less useful if they have a colloquially used label.  Additionally, it seems like we reach insurmountable impasses when people disagree on core intuitions about a concept.  Regardless, this essay is representative of my thought process when interacting with a text, in this case, Pritchard's book, Epistemology.  (It's a good book!  At least, apart from my problems with the overarching analytic project.)

Here's the essay.

Knowledge as a Form of True Belief;
Understanding as Forms of Justified Belief

Introduction

In this essay, I will be exploring the relationship between two forms of belief: knowledge and understanding.  Even though at their face these concepts might seem inseparable, I will show throughout the essay that knowledge is not a condition for understanding and vice versa.  Because they are not directly related in that way, discussing their relationship becomes mostly an act of explaining similarities and differences.  I will be primarily working with Duncan Pritchard’s conceptions of knowledge and understanding as laid out in his book Epistemology.  I do not raise issue with Pritchard’s intuitions about knowledge, but I scrutinize an aspect of his definition of understanding as cognitive achievement.  Pritchard argues that understanding is factive—that it requires that the belief being understood to be true.  I do not think this is as obvious as Pritchard claims, and disagree with him that this is necessarily the case.  I will argue that that there are actually two forms of understanding.  One form of understanding is factive as Pritchard says, and the other is not necessarily factive.  I will show why the best attempt to expand Pritchard’s definition of understanding as ‘strong cognitive achievement’ to include this second conception fails.  The not-necessarily-factive conception of understanding is characterized by the subject’s cognitive ability, while Pritchard’s conception is characterized by her cognitive achievement.  I will argue that the cognitive ability on display in both definitions of understanding is the subject’s ability to recognize that a set of related background beliefs form valid justification for her understanding.  I will compare knowledge and these two conceptions of understanding by looking at how they compare in terms of the four intuitions about knowledge.

Knowledge

I will first begin by explaining the main intuitions about knowledge, as described by Pritchard.  They are:

1.      Knowledge is a belief, which necessarily comes in the form of a proposition (Pritchard, 3).  A proposition shows a relation between subjects.  The belief “My roommate wears glasses” in standard propositional form for this might be Wears(my roommate, glasses).  This format will come in handy later when we examine understanding.

2.      The belief is true (Pritchard, 4).  A proposition is true if it matches what-is-the-case.  “My roommate wears glasses” is true if he actually does wear glasses.

3.      Anti-luck: the belief does not happen to be true largely because of luck (Pritchard, 5).  There are two main ways luck can prevent a belief from being knowledge:

a.       Intervening Epistemic Luck: An agent holds a true belief which in a normal environment is false, but in this case they got lucky by being in an abnormal environment that makes the belief true (Pritchard, 7).  For example, Steve sees a sheep-shaped-rock and believes that there is a sheep in the field.  It just so happens that there is a sheep in the field, hidden behind the rock, but that doesn’t make Steve’s true belief into knowledge.

b.      Environmental Epistemic Luck: An agent holds a true belief which in a normal environment is true, but because the environment is not normal, the agent has gotten lucky by basing their belief on the rare normal information (Pritchard, 8).  For example, Barney believes he sees a barn as he drives in his car, and he truly is seeing a barn.  However, because he is driving through a county where most of what look like barns are fake Hollywood-esque facades, he was just lucky in seeing the only true barn, and his true belief is not knowledge.

4.      Ability: the way the person comes to hold the belief is primarily attributable to the person’s cognitive ability (Pritchard, 6).  For example, they may perceive something directly, reason a conclusion from prior knowledge, or determine a reliable source to gain information from.  All of those rely heavily on the person’s cognitive ability.

The specific details about the anti-luck and ability intuitions take up chapters of space, so I won’t go into them here.  Luckily, those specifics are not necessary for this essay—the overview of the four intuitions is enough to explain the important similarities and distinctions between knowledge and understanding.

Understanding

I will now examine understanding.  Pritchard sticks to examining ‘atomistic’ cases of understanding rather than ‘holistic’ cases (Pritchard, 127).  Atomistic cases of understanding deal with a specific question.  For example, a tree falls in Susan’s yard.  An atomistic case of understanding would be whether Susan understands why the tree fell.  The atomistic case is reducible to a proposition.  “Why did the tree fall?” is asking “What is X that caused the tree to fall?” or Caused(X, fallen tree).  All atomistic cases of why or how something-is-the-case are reduced to a similar, specifically causal proposition (Pritchard, 127).

On the other hand, holistic cases are about understanding a broad topic (Pritchard, 128).  Susan's understanding of quantum computing is a case of holistic understanding. In this case, “quantum computing” is not easily reducible into a single propositional belief.  Instead, it seems to be made up of a bunch of propositions, which, if Susan generally understands them, means that Susan understands quantum computing.  We will stick to examining atomistic cases of understanding because they follow a propositional structure and are therefore more easily compared to knowledge.  Additionally, atomistic cases are important because they might be the key propositions that make up holistic cases, but that is not in the scope of this paper.

If we are looking at atomistic cases, let’s re-examine the case of Susan’s fallen tree.  Understanding why the tree fell seems to have to do with filling in the X in Caused(X, fallen tree).  Once we do so, we are left with a complete proposition which might look like: “Termites eating away at the wood caused the tree to fall” or Caused(termites, fallen tree).  Susan might believe this proposition, and it might cohere with her relevant true beliefs of how termites can weaken the wood of a tree.  However, Pritchard argues that understanding is factive.  That means that for Susan to understand why the tree fell, her belief of why it fell needs to be true.

CASE 1:
          The tree didn’t fall because of termites, but Susan believes it did.

Pritchard seems right to say that in Case 1 Susan does not understand why the tree fell.  Even if she has consistent relevant beliefs about termites and good justification for believing that termites caused the tree to fall, if her belief isn’t true then it doesn’t seem like Susan understands why the tree fell.  For that general reason, Pritchard claims that understanding is factive.

A Problem with Pritchard's Account of Understanding

However, this is an incomplete examination of understanding.  Even though Pritchard is right to claim that it doesn’t seem reasonable to say, “Susan understands why the tree fell,” it does seem reasonable to say that “Susan has an understanding of why the tree fell” even if her understanding doesn’t match up with what-is-the-case.  I will argue that we are looking at two forms of understanding.  The first sentence uses a factive definition of ‘understanding’.  The second sentence uses a definition of ‘understanding’ which is not necessarily factive.

This discrepancy isn’t handled in his book, so I brought it up with Pritchard and he proposed a solution that only uses his factive formulation of understanding. He claimed that in the reasonable sentence, Susan does have a true belief about something, just not what-is-the-case in the world.  Susan understands a theory for why the tree fell, but does not understand why the tree fell.  This seems to solve the problem without claiming that a second definition of understanding is at play.  However, under closer scrutiny, this explanation does not appear to hold.  Here is a criticism of the ‘true about a theory’ explanation for this discrepancy.

It helps to start by observing that such an explanation seems to imply that all causal beliefs now meet some sort of factive condition.  If a belief is not true about the world, then it will at least be true about a theory of the world.  There are two propositions being considered simultaneously:

P1:          Caused(termites, fallen tree)
P2:          StatesThat(Theory T, “termites caused the fallen tree”)

In Case 1, P1 is objectively false while P2 is objectively true.  Under Pritchard’s ‘true about a theory’ explanation, even though Susan believes both P1 and P2, because of the propositions’ objective truth values, Susan cannot understand P1 but does understand P2.  This is where I have a problem with Pritchard’s explanation.  I think that it can intuitively be said that Susan understands P1, that “termites caused the fallen tree”.  Even though P1 is not objectively true, Susan still understands it.  Her understanding is directly P1, without having to be re-aimed at a theory about P1, which is what Pritchard claims.

Pritchard’s single definition of understanding is not enough to account for the truths of both “Susan doesn’t understand why the tree fell” and “Susan understands termites caused the fallen tree”.  A cleaner explanation is to accept that there are two closely related definitions of understanding.  One of the definitions Pritchard accurately points out is factive, while the other is not necessarily factive.  “Susan doesn’t understand why the tree fell” is true because it uses the factive definition.  “Susan understands that termites caused the fallen tree” is true when using a different definition which doesn’t demand truth of the Susan’s understood belief.  The second statement would be false under Pritchard's factive definition of understanding.  For brevity, I will call the new and not-necessarily-factive conception of understanding: understanding* (understanding prime).

'Understanding Prime' as Justified Belief

Understanding* still requires more than belief.  Susan can’t be said to understand* that termites caused the fallen tree unless she also has sufficient background information about termites and trees.  So long as she believes additional propositions that describe why termites had the ability to cause the tree to fall, Susan might be able to understand* that “Termites caused the tree to fall”.  A possible set of background beliefs might be that “The tree had termites in it”, “Termites eat the inside of trees”, “Termites eating the insides of trees weakens them”, and “Sufficiently weakened trees will fall naturally because of gravity.”  Furthermore, Susan must also display the cognitive ability to realize that these background beliefs can be pieced together to form valid justification for the belief that “Termites caused the fallen tree”.  This ‘holding of background beliefs’ and ‘display of cognitive ability’ is necessary and sufficient for saying that Susan has understanding* in the not-necessarily-factive sense.  If Susan did not hold any background beliefs about termites being able to cause her tree to fall, or if she does not realize that her background beliefs form valid justification for the fallen tree, then it seems that Susan would not have understanding in any sense about “termites causing the fallen tree”.

Susan might not be consciously aware of this piecing-together thought process, but so long as it happens, she has understanding*.  This means that understanding* necessarily and characteristically requires Susan to be able to reflectively access justification for why the tree fell.  Not only can she say that the tree fell because of termites, she can justify it by saying that termites had the ability to cause that tree to fall.

‘Understanding prime’ puts emphasis on the self-reflective awareness of the understood* belief’s justification by specifying the type of cognitive ability necessary to have understanding*.  This description of cognitive ability is also a useful addition to Pritchard’s conception of factive understanding as ‘strong cognitive achievement’.

Pritchard’s description of understanding as a ‘strong cognitive achievement’ is essentially a slight modification to virtue epistemology.  Cognitive achievement is a cognitive success which is primarily attributable to a person’s cognitive ability (Pritchard, 132).  The ‘success’ of understanding is that the understood belief is true.  A cognitive achievement is ‘strong’ if either (a) the subject is overcoming a significant obstacle to their cognitive success or (b) if they display significant, skillful, relevant cognitive ability (Pritchard, 133).  However, Pritchard’s description of ‘strong cognitive achievement’ does not necessarily imply that the subject has self-reflective access to a justification for her belief, even if Pritchard argues that it practically does imply this.  Pritchard argues that the ‘strong’ condition essentially means that the agent must display the cognitive ability to overcome the significant obstacle of cognitively piecing together the relevant pieces of information (Pritchard, 134).  If piecing together the information is not a significant obstacle, this is because the subject is displaying significant cognitive skill; piecing together the information is easy because she is skilled (Pritchard, 134).  As we saw in the description of understanding prime, this cognitive ability of ‘piecing together background beliefs’ means that the subject realizes that the background belief(s) form a valid justification for her understood belief.  Although the definition of ‘strong cognitive achievement’ does not necessitate the cognitive ability of understanding prime, reflectively available justification does seem inescapable in cases of ‘strong cognitive achievement’.

I think it would be best for ‘strong cognitive achievement’ to specify that ‘strong’ means the subject must display cognitive ability as described in ‘understanding prime’.  This would barely change the factive definition of understanding.  However, it has the benefit of necessitating that the subject has reflective access to justification of their understood belief.  It seems intuitive that for somebody to understand a causal belief, they must have reflectively available justification for that causal belief.

Comparing Knowledge, Factive Understanding, and Understanding Prime

Now that we have described knowledge and two forms of understanding, how are they similar and different?  Let’s look at the four intuitions about knowledge and see how factive understanding and understanding prime match up.

1.      Knowledge, factive understanding, and understanding prime all place additional conditions onto a held belief.  One difference is that the forms of understanding we looked at specifically deal with causal beliefs, while knowledge is not restricted in such a way.

2.      Both knowledge and factive understanding require the belief to be true, but understanding prime does not require the understood belief to be true.

3.      Anti-Luck:

a.       Knowledge doesn’t occur in both forms of lucky true belief.

b.      Factive understanding cannot occur in cases of intervening luck because the success is primarily attributable to the luck, not the cognitive ability.  However, factive understanding can occur in cases of environmental luck.  For example, Barney uses sufficient cognitive ability to determine that he is looking at a barn.

c.       Understanding prime is not affected by the anti-luck condition because the anti-luck condition is about luckily being true, and understanding prime does not require truth.

4.      Ability:

a.       Knowledge requires some form of cognitive ability.

b.      Both factive understanding and understanding prime require a specific type of cognitive ability: the cognitive ability to realize that background beliefs form a valid justification for the understood belief.

As a final comparison, knowledge does not necessarily have the availability of self-reflective justification which understanding necessarily possesses.

Works Cited

Pritchard, Duncan (2016, Jan 21). Epistemology.

post last updated 2017, June 30

changed: shortened preface