Non-epistemic

On the internal realist view there cannot be a mismatch. The real structures are the structures our recognition abilities and justification conditions identify. The real boundaries coincide with the boundaries we draw, since the former are identical with the latter. Something is a dog, if the sentence This is a dog is justified when it is applied to it, for the very kind DOG owes its existence to the justification conditions. As a result, at least some of our sentences are bound to be true, namely, those simple sentences whose justification conditions are responsible for drawing the boundaries. So truth cannot completely diverge from justification. There are sentences for which to be true is to be justified. Consequently, there is a conceptual connection between truth and justification. The connection is not purely contingent. Truth is then not completely non-epistemic. 

The correspondence theory of truth in the stipulated sense of the previous chapter is incompatible with this account on two counts. The correspondence theory in the stipulated sense is a Tarskian theory with reference understood as a non-epistemic relation to entities that are mind-independent in the sense of (MR1). The present account considers reference an epistemic relation, since it hinges on justification conditions, which are clearly epistemic. It also fits badly with the idea that the entities words refer to are ontologically independent of the human mind. The disagreement concerns both the nature of the reference relation but also one of the relata. This latter point may be less clear, since the three-step recipe does not say anything about the ontological status of the entities we refer to. So why does it naturally tie in with (IR1) rather than with (MR1) ... 

Besides, metaphysical realism does not fare any better in practice. In real life, the question concerns the adequacy of particular conceptual schemes. Suppose there are two competing conceptual schemes in a given domain. Which one is right The idea of a mind-independent structure is of little help here. The advocates of the two schemes claim that it is their own scheme that captures that structure. But they have only epistemic criteria to substantiate their claims. Adequacy - in the metaphysical realist view - is completely non-epistemic, so there might be no way of knowing which scheme is adequate. To say that one of the schemes is adequate but we do not know which one leaves the situation completely symmetrical. To put it differently,... 

Once we accept that the plausibility of (A) derives from (B), it becomes clear that the assumption is very much in the spirit of internal realism. Internal realism regards reference is an epistemically loaded notion, since it argues that it is fixed by justification conditions. Metaphysical realism, on the other hand, treats reference as non-epistemic. If it did not, it would have to deny (MR2), the thesis that truth is radically non-epistemic. Since truth is determined by reference, were reference epistemic, truth would become epistemically tainted. 

More generally, in modeling environments, it is often needed to accept certain assumptions or simplifications. The kinds of assumptions and simplifications made (conservative or optimistic) ultimately rely on non-epistemic value judgments rather than on epistemic justification. 

An important distinction exists between belief-based attitudes (B, K, OU, EU, DB) and acceptance-based attitudes (A, AEB). This involves the role of non-epistemic values as justification for the type of attitude taken to a proposition. In the belief-based attitudes, the value judgments of a subject are irrelevant because whether or not one takes something to be true does not depend on whether one sees value in taking it to be true. In contrast, in acceptance-based attitudes, values do play a role as supporting justification. This is related to the attitude of acceptance more being tied to a particular action in a particular context, rather than aiming at the discovery of a true value. 

Concerning degrees of behef (subjective probabilities), these may or may not involve values depending on what types of justification (epistemic vs non-epistemic) are invoked as evidential statements on which the degree of belief is based. If these evidential statements only concern attitudes of knowledge, behefs or evidential uncertainty, non-epistemic values are not present in the... 

Furthermore, there often is a form of evidential uncertainty for the assessor in relation to evidential propositions. Finally, it is possible that the assessor has accepted data, models or judgments the justification of which is (partly) based on values. Hence, the justification may not only of Epistemic nature (Ej), but also may involve of Non-Epistemic nature (NEJ). Hence, the risk description is not necessarily value-free, depending on the nature of the reasons because of which a particular propositional attitude is adopted regarding the evidences. A qualitative assessment of evidential biases can make such value-laden considerations explicit. 

Apart from the probabilistic risk measurement, a sensitivity analysis is made, identifying the elements in the model to which the risk measure is most sensitive. For those probabilities to which the model is most sensitive, a qualitative assessment of the strength of justification and the direction of bias is made. Finally, the sensitivities, strengths of justification and (when relevant) the direction of biases are grouped in a set of matrices, showing the overall picture of sensitive model elements and their epistemic and non-epistemic justification. 

A proposed evidence assessment scheme for the evidential support for a subjective probability P,(A I EJ,NEJ) is presented in Table 3, accounting both for epistemic and non-epistemic justification. The scheme may be used to make explicit what type of evidence the assigned probability value is based on, as well as how strong that evidence is believed to be. When relevant, the direction of bias inherent in the evidence can be judged by the assessor. 

On the top-right corner, the various risk assessment model elements, i.e. the various subjective probabilities in the risk model, are identified. These are subsequently mapped according to their sensitivity values and their aggregate score in terms of epistemic and non-epistemic justification in a set of matrices. 

The difference between the epistemic model of explanation and the ontic one becomes clear at this point Cause... because, as Philip Kitcher elegantly puts it, is the slogan of the epistemic school. Explanations frequently refer to causes simply because they explain much, but there is no conditio sine qua non (1989). If it turned out that other non-causal regularities were better explanatory vehicles the causal explanation would be dropped. The adherent of the ontic model of explanation, on the other hand, maintains that any valid explanation has to cite causes of the explanandum because explanation owes its only possible sense to a thorough assumption about the ontic constitution of our world. Those who want to capture the validity of explanations in their dependency on causes have to give an answer to the question of what causality consists in. Here, I sketch the two major accounts of causality. 

If one claims that analysis destroys the form of the substance and that synthesis restores this form, this does not suffice to tell us what this form actually is. If, on the other hand, substantial form is something non-physical that disappears and then reappears, then this does not explain how a chemical procedure can affect a non-physical form . If substantial form simply means the substance itself, then we end up with a circular explanation The substance has been altered because the substance has been altered. Any of these three choices leaves us epistemically unsatisfied. To be fair, there is another, more interesting way of understanding substantial form that serves as a precursor to Boyle s mechanistic structuralism. Substantial form can be understood as meaning the inner structure that gives the substance its essential properties. The problem with this understanding of substantial form, however, is that it stUl does not teU us what this inner structure is. Boyle latches on to this idea of inner structure but seeks to understand it in strictly mechanistic and corpuscularian terms, without any reference to substantial form. 

The notion of academic drift as we use it here is meant to refer to the set of phenomena described by Martin Trow as cited above. Academic drift may be seen as corresponding to what the Australian scholar Malcolm Skillbeck has alternatively called academic creep (Skillbeck 2003, p. 5) and to a certain extent to what the Dutch scholar Aant Elzinga has called epistemic drift (Elzinga 1985). In this chapter, however, we prefer to stick to the notion of academic drift as this is the standard use in the literature that we are reviewing. In the remainder of this chapter academic drift in professional non-university engineering education is therefore understood as follows. 

Each turbine has the key subassemblies shown in Figure 4 and these represent the subassemblies likely to have large or moderate epistemic uncertainties in their reliability estimates. We also create a further generic subassembly category that collectively represents the other non-key subassemblies for which it is anticipated to have small epistemic... 

Kaplan (1997) proposes the so-called probability of frequency approach to risk assessment, based on a risk concept in line with risk definition C6 (R = P C), where subjective probabilities are used to express uncertainty about true frequen-tist probabilities. The assessment thus focuses on quantifying uncertainty about an underlying true risk, which is estimated. Kaplan s view is strongly tied to realism, as the risk description focuses on a true risk as determined by experts. Closely related perspectives are those where uncertainty is quantified around a true risk, such as in the traditional Bayesian perspective where uncertainty is quantified in relation to model parameters (Aven Heide 2009). Such uncertainty quantification can also be done using non-probabilistic representations of epistemic uncertainty (Helton Johnson 2011). These methods typicdly consider a risk problem in a highly mathematized form. 

Reduction, explanation, and the like are epistemic activities, and the mere fact that such equivalence or biconditionals exist is no guarantee that they are, or will ever become, available for reductive or explanatory uses . J. Kim, Concepts of Supervenience , p. 173. It is interesting to note, however, that Kim no longer thinks that supervenience supports the non-reductionist program. Cf. his Mechanism, Purpose, and Explanatory Exclusion , in J. Tomberlin (1989) (ed.), Philosophical Perspectives, Vol. 3 (Atascadero, Calif. Ridgeview), 77-108. 

---