• 7.1: Evidence Neutrality
• 7.2: Argument Against Intuitions
• 7.3: Judgment Skepticism
• 7.4: The Judgment Skeptic’s Mistake
• 7.5.1: Traditional Skepticism
• 7.5.2: More on Evidence Neutrality
• 7.6: Epistemic Conservatism
• 7.7: Reflective Equilibrium

    A move an equivalency theorist might make is to claim evidence justifies belief with regard to truth-directedness. A belief p is justified in relation to normative reasons for holding p. From a third-person perspective only reasons that are true will justify belief. Such reasons align with evidence because their justificatory power derives, in part, from their factivity. This, however, is not a move an equivalency theorist will be comfortable with. There are also subjective reasons, and these reasons are not always factive. Such a theorist posits a symmetry thesis:

• Symmetry: Normative reasons and evidence are symmetrical with regard to subjectivity and objectivity.

According to Symmetry one can speak of subjective reasons and subjective evidence along with objective reasons and objective evidence. What is objective is always factive and what is subjective is not always factive (or, is non-factive). One problem with this neat division is that cases (like the Coop-Petrol case) apply pressure to this divide. It seems that normative reasons must always be factive. Littlejohn (2010) mentions that Fantl and McGrath (2009) are committed to saying:

(1)   What justified Coop in giving Audrey the toxic stuff was that there was gin and tonic in that glass.

And, (1) entails (3):

(3)   Coop was justified in giving Audrey the toxic stuff (in part) because there was gin in the glass.

A factive theorist can point out that (1) entailing (3) is only a problem if p is true and q is true. That is, Coop was only justified in his reason for acting as he did if there was gin, and not petrol, in the drink he gave Audrey. However, there was not gin in the drink, so it’s not true that ‘p because q’ is true. Thus, Coop was not justified in his action. Fantl and McGrath try to block (1) entailing (3) by resorting to subjectively-based (2):

(2)   What justified Coop in giving Audrey the toxic stuff was that there was gin and tonic in that glass, as he thought at the time.

I agree with Littlejohn that (2) doesn’t block the entailment. That Coop thought that the clear stuff was gin and tonic doesn’t prevent the implication in (1) that there was gin and tonic in the glass and this reason justified Coop in acting as he did. So, it does not block (1) from entailing (3). If one takes Littlejohn’s suggestion that normative reasons are factive, the next question becomes whether evidential beliefs must be true. Timothy Williamson thinks evidential beliefs must be true, whereas Jim Joyce (2004) argues against this idea.

    For Williamson, “If e is evidence for h, then e is true” (2000: 201). If something is inconsistent with the facts, then it’s not evidence because, “No true proposition is inconsistent with my evidence, although I may think that it is” (2000: 201). According to Joyce, Williamson holds his factivity view because he conflates subjective and objective senses of justification. Williamson is only focusing on the objective (third-person) form of justification, so it’s no wonder that beliefs justified by evidence must always be true. Evidence is always factive for Williamson, so that which it justifies must also be factive.

    Joyce, like Fantl and McGrath, ask us to consider the subjective sense of ‘justifies’ in terms of a subject’s reasons that she possesses for a belief. These justifying reasons need not always be factive, and, as Joyce points out, an agent will only mistakenly consider evidence justifying a belief if there’s a false belief present. For instance, this occurs when a juror holds the false belief that a witness is reliable and believes based on that false belief that the witness’ testimony is evidence for the innocence of the defendant even though the witness is lying.

    I’ll close by pointing out a couple of assumptions underlying the Joyce and Fantl-McGrath response. These assumptions might be undermined as a way of arguing for the factivity of normative reasons and evidence.

• Assumption 1 (AS1): ‘Objective’ goes with ‘factive’, whereas ‘subjective’ goes with ‘non-factive’.

(AS1) implies that ‘objective’ is agent-neutral (applying to external facts), whereas ‘subjective’ is agent-relative. Another way of understanding ‘objective’ is in terms of ‘abstraction’. Even if something is abstracted away from an agent it may none-the-less be relative to an epistemic situation and by implication be relative to the mental state of a (hypothetical) agent. So, ‘objective’ can still be agent-relative in that sense. It might be that an actual agent can inhabit such an epistemic state. What matters is that an actual agent is not required to inhabit such a state for there to be evidence or reasons.

• Assumption 2 (AS2): An agent only has (possesses) evidence or reasons subjectively. Objective evidence (or reasons) is evidence there is, where subjective evidence is evidence an agent has.

Contrary to (AS2) it’s possible to argue that subjective access to evidence, through armchair reflection, is not required for evidence possession. If evidence is factive, then an agent may possess such evidence even though an agent is not subjectively able to recall or grasp that evidence.

The distinction is: provide vs. consist. Williamson uses this distinction to mention an objection to his view that all evidence is propositional. For Williamson even perceptual experiences, which are often regarded as non-propositional evidence, consist of propositions. An objector might claim: “Experiences provide evidence; they do not consist of propositions” (197, italics mine). However, only propositions we grasp can be used in confirmation, inference to the best explanation, and choice between rival hypotheses. Even though words fail to completely capture perceptual experience it does not mean evidence is non-propositional. Instead, experience makes propositions e₁…e_n count as evidence for a hypothesis h. Having an experience bestows the status of evidence on propositions. As such, evidence is inextricably linked to (and mediated by) propositions. Thus, experience consists of propositions.

It is possible to ask the same thing of epistemic reasons and evidence: Are they inextricably linked? There are defenders of two theses concerning this question:

• Inseparable:  Where you find one you find the other (i.e., reasons and evidence serve the same function, appear under the same analysis, or are constitutionally equivalent).
  • For every proposition p, if p is a reason R then p is evidence E.
• Separable: Reasons and evidence come apart (i.e., in some scenarios you have reasons but no evidence, and vice versa).
  • There is some p such that p is an R but not an E.

Now I’ll relate this to the provide/consist distinction. One way of arguing for Inseparable is by claiming ‘having’ evidence for the truth of p ‘provides’ you with an epistemic reason for believing p. In response one might argue for Separable by showing evidence for p doesn’t always generate a reason to believe p. A strategy to counter this move is to claim the evidence for p is not really (good) evidence for p. What is taken as evidence for p doesn’t ‘consist’ of evidence; it doesn’t have the status of evidence because its status is undercut by other pieces of evidence.  These moves have the following assumptions:

• Pro-Inseparable: If you have good evidence for p, then you have a good epistemic reason for believing p.
• Pro-Separable: Rejects the assumption endorsed by Pro-Inseparable.

Pro-Inseparable claims ‘providing’ sanctions ‘believing’. ‘Believing’ connects to ‘consisting’ in that one is ‘believing’ appropriately if that believing is based on that which has the status (consists) of good evidence (reasons). By transitivity ‘providing’ sanctions ‘consisting’.  A way to argue for Pro-Separable is to show you can have good evidence for p without that evidence grounding a good reason for believing p because the reason is not based on the evidence. As a result, it’s not the case where you have good evidence you always have a good epistemic reason. The epistemic reason needs to be appropriately linked to the evidence to result in ‘believing’ in a way that’s sufficient for the belief to be justified. Simply claiming the two entities are inextricably linked (i.e., where you find one you find the other) doesn’t secure this connection. There’s another way of putting this point.

‘Providing’ focuses on the function of evidence or how it’s used in an argument. Williamson takes this line by arguing evidence ‘is’ (consists) only in so far as it ‘functions’ (provides). For evidence (experience) to play its evidentiary role within an argument it must be propositional. Because experience functionally provides evidence for hypotheses, and evidence must be propositionally grasped in order to be used, experience consists of propositions. Williamson argues for ‘consists’ by way of ‘provides’. The problem with this is that something may ‘consist’ without ‘providing’. I can be justified in believing that p even if no agent has engaged in the activity (function) of justification. This is because p’s status as evidence justifies believing in p in a way that doesn’t depend on anyone having used it in argumentation. It doesn’t require that the evidence is possessed, grasped, or used. That it can only ‘function’ a certain way if it ‘is’ a certain way simply shows that ‘function’ (provide) depends on ‘status’ (consist). It doesn’t show that ‘consist’ can be derived from ‘provide’ when it comes to evidence.

ABSTRACT. The Requirement of Total Evidence (RTE) asks an agent to make her confidence in a belief proportional to the support it receives from her total evidence. This paper examines (RTE) as a norm of epistemic rationality and argues that it is problematic. Looking at the work of Peter Achinstein (2001) on the notion of evidence it becomes clear that (RTE) endorses a view of the constitution of evidence that is neither necessary nor sufficient for something to count as evidence. To overcome this and other deficiencies associated with (RTE) a move is made to an objective view of evidence. This move aligns epistemic rationality with scientific rationality in seeking to capture veridical evidence. It also leads to a new norm of epistemic rationality—the Proper Subset Evidence Requirement (PSER).

Click on the following link to access a presentation on the paper.

(RIPD-T) The reliability of a testimonial assertion is inversely proportional to the distance of that assertion from the original testimonial assertion.

The Reliability Inversely Proportional to Distance principle (RIPD-T) accounts for the value of eye-witness testimony. If a person actually witnessed an event, then (ceteris paribus) their testimony is typically more reliable than, say, someone who heard about the event from another person. Setting aside cognitive biases and the reliability of agents according to the reliability of the senses (i.e., whether someone is hard of hearing, has poor eyesight, etc.) a problem arises.

   There are cases where the original testimony is less strong than subsequent testimony based on the original assertion. Such is the case when an event does not cognitively register, but subsequent assertions of the testimony better approximate the truth. There are historical examples where native people, upon seeing ships off the coast, either did not see the ships or did not see the ships as ships. A native person who is an eyewitness to a ship, which seeks to conquer the person’s native land, may initially report that they saw a large, sculpted floating piece of wood today. As the testimony is transmitted throughout the tribe it may be refined to a floating boat, a floating boat with people, until it reaches the Chief of the tribe and it is reported as a floating boat full of foreigners. This happens in everyday life when things are new and one does not yet possess the correct concept or language required to approximate the truth in testimony. Subsequent assertions of the testimony, especially in a dialectical context, can result in positive refinement of the assertion. This leads to a principle contrary to (RIPD-T) called the Reliability Proportional to Distance principle:

(RPD-T) The reliability of a testimonial assertion is proportional to the distance of that assertion from the original testimonial assertion.

The (RPD-T) principle holds that testimony becomes reliable in proportion to its distance from the original assertion. I mention these principles to bring out a point about evidence.

   Testimony is often regarded as evidence, or as a source of evidence. Legal contexts assume (RIPD-T). In addition, philosophical accounts of testimony, like Robert Audi’s, associate testimony with non-inferential knowledge. One way to support such an account is by using an (RIPD) principle, namely, the closer you get to an experience, unimpeded by inferences that can go astray (probabilistically), the more veridical the testimony. Some accounts of evidence (like an account endorsed by Clayton Littlejohn here and in other places) is that p is part of your evidence-set if you non-inferentially know p. I wonder if these accounts of evidence implicitly endorse something like (RIPD) with regard to the strength of evidence. If so, the following principle is possible:

(RIPD-E) The strength of a piece of evidence is inversely proportional to the distance of that evidence from the experience (fact) that grounds that evidence.

Again, this seems to work in a legal context. If the fact is the evidence, as in the case of physical evidence presented in court, then that evidence is regarded (ceteris paribus) as stronger than a report about the physical evidence, which can leave out details and otherwise distort characterization of the evidence. But, an (RPD) principle is possible to formulate, even in a legal context. A jury may not know what they are looking at or what is relevant about a piece of physical evidence until it is explained. However, that explanation about the evidence may involve an interpretation of the evidence. Some interpretations of the evidence and its relevance to the guilt of the defendant may be more accurate than others. It is possible that given enough subsequent refinements of reports on the evidence, perhaps refinements that fill out the factual details, that the evidence becomes stronger in the eyes of the jury. In such a case, an evidence-specific (RPD) principle is possible:

(RPD-E) The reliability of a piece of evidence is proportional to the distance of that evidence (or propositional report on the evidence) from the original physical evidence, experience, event, or evidential proposition.

The proportional and inversely proportional principles pull in opposite directions. Philosophers whose account of testimony or evidence implicitly assume (RIPD) need to hang up the telephone (couldn’t resist) and sort out how an (RPD)-type principle can be explained away, undercut, or otherwise discharged. Otherwise, it seems reasonable to disconnect veridicality from closeness to the original bit of testimony or evidence.

Looking at the lineup, hopefully the talks on Saturday target the concept of ’evidence’. The topics relate to evidence, but they are more meta-topics that range over evidence. I’m sure Neta, Goldman, and Siegel will target ‘evidence’ as they discuss, for example, higher-order reasons, reliabilism and evidentialism, and cognitive penetration. Either way, this should be a great conference. You can apply to attend the conference from the conference website. Unfortunately, I’m unable to attend the conference, but the conference organizer (Yuri Cath) mentioned here that there might be audio from the conference. I will keep you informed if a link to the audio surfaces.

1. Evidential justification is distinct from non-evidential justification.
   • Df: Evidential justifiers (e.g., other justified beliefs) are capable of justifying beliefs by serving as evidence for those beliefs, whereas non-evidential justifiers (e.g., reliability, coherence) are things in virtue of which  beliefs are justified (i.e., they are not evidence for the beliefs but that on which justification supervenes, for example).
2. The Sellarsian dilemma must be restricted to evidential justification; otherwise, it is a nonstarter.
3. The distinction between evidential and non-evidential justifiers makes it possible to hold separate the Belief Principle and the Grounds Principle.  
   • Df: The Belief Principle claims that only beliefs can evidentially justify beliefs, whereas the Grounds Principle claims that all justified beliefs have grounds (i.e., evidential justifiers).
4. Doxasticism is the conjunction of the Belief Principle and the Grounds Principle.
5. When the Sellarsian dilemma is restricted to evidential justification (as premise 2 indicates it must be), it is only an argument for the Belief Principle.
6. Thus, the Sellarsian dilemma is not an argument for coherentism (doxasticism); such an argument requires an argument for the Grounds Principle, which the dilemma does not provide.
7. However, doxasticism can be rejected on independent grounds (assumption not argued for, assumed from the literature).
8. Rejecting doxasticism for independent reasons results in rejecting the Belief Principle and the Grounds Principle.
9. Thus, in rejecting doxasticism, the Sellarsian argument for the Belief Principle turns out to be an argument for rejecting the Grounds Principle.
10. Therefore, the Sellarsian argument is an argument for non-evidentialism (i.e., not all beliefs must have evidential justifiers as grounds), and beliefs can be justified by factors external to the agent (i.e., externalism is true).

In laying bare Lyons’ argument we have already accomplished something because it’s easy to get lost in the mounting of distinctions, positions, and counter-positions in his paper. For a defense of each premise I recommend reading the whole paper, but for the remainder of this post I’ll focus on one aspect of the argument. This aspect might undermine the entire argument.

Lyons responds to an objection to his argument. The objection is that he’s working with too narrow of an understanding of evidence and this begs the question against non-doxastic evidentialist theories. Notice that the argument above is restricted to a brand of doxastic evidentialism. Perhaps a different understanding of evidence, one that included, for example, physical evidence like a murder weapon, would not commit one to doxasticism. As a result, it might be possible to satisfy evidential justification without being led into the Sellarsian dilemma. Lyons responds to this charge without argument, by citing philosophical tradition. Put candidly: if you’ve got a problem with my usage of evidence, it’s their problem not mine! The people he’s referring to are Feldman and Conee, Haack, and Alston. Lyons says that it’s their conception of evidence that implies the Belief principle, and he’s merely borrowing this conception for the sake of argument. Even if we grant Lyons that his use of the notion of evidence is not question-begging against his opponents, it ‘s still problematic for his argument.

There’s solid motivation to move away from a doxastic notion of evidence. If only beliefs can serve as evidential justifiers, even for the sake of argument, then the notion of justification hinges on defending and substantiating a doxastic notion of evidence. Recently, Ram Neta (in “What Evidence Do You Have?”) has provided extensive counterexamples to a doxastic notion of evidence as cashed out in its various theoretical forms (e.g., reliability, JTB, E = K, coherence, deontology, and so on).

As it turns out, Lyons makes a double-mistake. First, he fails to argue against doxastic theories, and merely cites that the literature supports the idea that “doxasticism is a lost cause,” thereby motivating premise 7 without argument. Second, he bases his argument on doxasticism because it is a notion endorsed by mentalism and other internalist-based epistemologies. Without arguing against a non-doxastic understanding of evidence his argument is something of a straw-man. That is, he set-up an easy opponent to knock down. To sustain his argument Lyons must discharge the burden in the literature which points away from a doxastic conception of evidence in the first place, a burden that undermines the initial premises (assumptions) in his argument unless it can be discharged.

A problem with the truth-maker view of facts is the existence of negative truths. If there always must be something that makes true that which is true, what are we to say when there is no entity in the world that makes a negative truth true? What entity “makes true” propositions like {There is no God} or {Carbon monoxide is an odorless gas}? What non-occurrence event of the existence of God makes it true that {There is no God}? Does the non-existence of the property of odor in carbon monoxide make true the proposition that {Carbon monoxide is an odorless gas}? How is lacking a property able to “make true” when the basic truth-maker principle requires there to exist at least one entity that necessitates the truth of that which it makes true? Is odorlessness an entity capable of such a task?  

This problem is relevant to the concept of evidence. The nonoccurence of events is said to provide negative evidence. Such evidence is used in inferential contexts to draw certain conclusions. Consider the Sherlock Holmes’ mystery Silver Blaze. At night someone stole the horse Silver Blaze from the stable where he was kept. A watch dog and two grooms also lived in the stable. The grooms said they heard nothing during the night. Consider the following propositions:

• b: The dog is a watch dog that would bark if someone attempted to steal Silver Blaze, and this would wake the grooms and alert them of the intruder.
• e: The dog did nothing during the night.
• h: The person who stole Silver Blaze knew the dog well.

The probability of the hypothesis given the evidence and background information leads Holmes to conclude P(h|e & b) = High. The incident that allowed Holmes to conclude that h is likely was the non-occurrence of barking by the dog during the night. How does the state of affairs of non-barking (as an entity) make true (or probable) the hypothesis? If negative evidence is admissible as facts in the truth-maker sense where does this lead? Does the non-occurrence of my death (or bodily injury) while driving “make true” the proposition that {Driving is safe}? After all, the incident of non-occurrence of injury does increase the probability of the safety of driving. One problem is the scope of such commitments. Would all non-occurrence events equally “make true” propositions true? Inferentially speaking, would negative evidence, as “making true”, lead to the irrational assignment of probabilities in various hypotheses?

Imagine one dumps truth-maker theory and opts for truth-bearer theory instead. This has the upshot that propositions can be properly modeled and made to “obey” the rules of probability. Facts simply bear truth, they don’t make it. But, this leads one to wonder where the truth that propositions come to bear originates? If not in states of affairs, then is their truth a matter of mere stipulation? To be a fact is to be a true proposition. This can be stipulated as the intersection of the set of all propositions that exist in a world w and the set of all propositions that are true in w. But doesn’t the existence of physical evidence (or non-inferential evidence) tell against such a view? Why would one need to report on the evidence or duplicate the evidence in a that-clause when the evidence is before one’s eyes? This seems especially troubling when producing the evidence in court has greater probative value (e.g., think of O.J. trying on the leather glove during his trial) than a that-clause about the same evidence. If the claim is that all evidence must be rendered propositionally, then producing the evidence in court should not affect the probative value over and above the value when the evidence is propositionally presented. For instance, that {The glove doesn’t fit O.J.}, relative to certain background information, should have the same effect as O.J. trying on the glove in court and the jury seeing it doesn’t fit. Yet, perceptually viewing the material evidence in a state of affairs does seem to affect the probative value of the evidence.

For me, the jury is still out on a truth-maker versus a truth-bearer view of facts. Perhaps a hybrid view is possible. On such a view, facts might be rendered as either truth-makers or truth-bearers depending on the context of confirmation. Different contexts impose different standards of evidence.  For example, testimonial evidence carries different probative value depending on whether one is in the context of science or law. To this end, coming up with principles that guide the rendering of facts to various contexts of justification seems a worthwhile pursuit.

• Are facts truth makers or simply truth bearers?
• How does a propositional account of evidence accomodate non-inferential evidence?
• What is the relationship between epistemic and doxastic justification?
• What theory of the epistemic basing relation is most tenable (i.e., causal, counterfactual, doxastic, causal-doxastic)?
• How can the same epistemic reason be both normative and explanatory?
• How does my principle (evidence and reasons for belief–ERB) result in epistemic justification when all things are considered (e.g., all evidence and reasons are accounted for)?
• How is ERB a more defeasible principle of epistemic justification than its competitors?
• If evidence is a subset of the total facts about a case is one still rationally required to assess the total evidence in one’s evidence set when assessing the justification of a belief? If the total evidence principle is not rationally required, then what is the alternative principle that prevents irrational yet justified beliefs?
• What are my responses to the three arguments against foundationalism proposed by Howard-Snyder and Coffman (2006)?
• How is my version of foundationalism different from Alston’s two-tier model?
• How is my reasons and evidence-based theory of justification related to evidentialism of the Conee and Feldman type?
• How does my factive account of evidence defend itself against evidence is sometimes non-factive views?
• Does my account of justification lend itself to an account of knowledge? How does it handle the sort of counterexamples proposed by Neta in “Defeating the Dogma of Defeasibility” found in the book on Williamson’s theory of knowledge?
• How does my account of knowledge relate to or differ from the E = K thesis?

Here’s to the New Year!

I close by drawing your attention to several recent discussions about evidence: here, here, here, and here.