Tải bản đầy đủ - 0 (trang)
4 Is a Purely Descriptive Axiological Turn Feasible?

4 Is a Purely Descriptive Axiological Turn Feasible?

Tải bản đầy đủ - 0trang


P. Kroes and A.W.M. Meijers

of technology. Is it possible to take a purely descriptive stance towards engineering

practice? The answer to this question depends on whether as a matter of principle or

as a matter of fact, a sharp distinction between descriptive and normative claims can

be maintained. Whereas in our empirical turn paper the distinction between the

descriptive (synthetic) and conceptual (analytic) played a crucial role, now the distinction between the descriptive and normative is at stake. This distinction has been

challenged in the debate about thin and thick ethical concepts.

The notion of a thick ethical concept was introduced by Williams (1985) in his

critique of fact-value theorists who “are bringing their distinction to language rather

than finding it there and, in addition, are unreasonably expecting that when the distinction is revealed it will be found very near the surface of language” (p. 130).

Williams observes that many concepts, such as treachery, promise, brutality and

courage, are neither just descriptive nor just prescriptive/evaluative; their application is determined by a combination of fact and value. He calls them ‘thick’ ethical

concepts. Statements using thick ethical concepts do not fit neatly into the pigeon

holes of factual (descriptive) and evaluative (prescriptive) statements. They have

features of both. Somebody may, given certain circumstances, be called rightly or

wrongly courageous but this expresses a value judgment in addition to a factual

statement. When people disagree about whether somebody behaved courageously

or not, they may try to resolve their disagreement by analyzing more closely the

facts of the matter, that is, the person’s behavior under the given circumstances. But

they may also try to resolve their disagreement by analyzing and comparing more

closely their normative standards for courageous behavior in this case. With regard

to the use of thick concepts like courageous, however, there is no guarantee that a

recourse to either the facts only or to the normative standards only will be able to

resolve the disagreement. According to Williams the application of thick concepts

“is at the same time world-guided and action-guiding” (p. 141); they are both

descriptive and prescriptive. In contrast to thick ethical concepts, thin ethical concepts, such as good or right, lack any or almost all factual content.9

The problem with regard to thick ethical concepts is how to interpret the role of

facts and values in determining their meaning. Williams questions a particular

account according to which the application of a thick concept is only determined by

its descriptive elements; so it assumes that evaluative elements play no role in this.

In other words, for any thick concept “you could produce another that picked out

just the same features of the world but worked simply as a descriptive concept, lacking any prescriptive or evaluative force” (p. 141). According to this account a thick

ethical concept is, therefore, simply a descriptive concept on top of which an evaluative element is added. According to this so-called strong separationist line of

thought it is possible to disentangle the facts and values involved in thick


Williams doubts that it is always possible to come up with a descriptive concept

that captures the descriptive content of a thick concept. In the example above, about

the disagreement about calling somebody courageous, it may be assumed more or


Williams himself does not use the notion of a thin ethical concept.


Toward an Axiological Turn in the Philosophy of Technology


less tacitly that it would be possible to separate the relevant facts involved in calling

somebody courageous. If that would be possible then indeed there are in principle

two ways to interpret the disagreement: either there is disagreement about the relevant facts involved or about the moral norms involved (or about both). On a nonseparationist account the disagreement is of a different nature; it concerns a unitary

whole of facts and values that cannot be disentangled into its factual and evaluative


How to interpret thick concepts and whether or not they are a kind of concepts

sui generis, different from factual and evaluative concepts, has become a matter of

debate (Kirchin 2013). For our purposes it will not be necessary to enter into this

debate. More important for us is that as a matter of fact thick concepts, as Williams

suggests, are part and parcel of the surface structure of ordinary language and what

this implies for the possibility of a purely descriptive axiological turn. First we will

have a brief look at the use of thick concepts in engineering practice.

Given that engineering practice is a thoroughly value-laden practice it comes as

no surprise to observe that it is so to speak loaded with thick concepts. Key concepts

such as safe, dangerous, efficient, wasteful, reliable, user-friendly, environmentfriendly, sustainable, flexible etc. all appear to be thick concepts.10 On the assumption that these concepts have been operationalized into clear, objective measurable

lists of specifications, these concepts may be taken, prima facie, to lead to factual

statements: “X is safe to use for doing Y”, for instance, then may amount to the

factual claim that X satisfies a specific list of (measurable) criteria when used for

doing Y. Nevertheless, even in those circumstances, there appears to be an evaluative element involved in the claim, namely an (implicit) recommendation to use X

when one wants to do Y. Moreover, the acceptance of the operationalization itself

implies a value judgment to the effect that this operationalization of the notion of

safety is (morally) acceptable. Thus, although the application of thick engineering

concepts in particular circumstances is clearly world-guided, it involves at the same

time value judgments (i.e. is world-guiding), even if these concepts have been operationalized in terms of objective measurable criteria.

What implications may be drawn for our distinction between a descriptive and

normative axiological turn from Williams’ distinction between thick and thin concepts and from the fact that thick concepts are ubiquitous in engineering practices?

To begin with the latter, the fact that the use of thick concepts is part and parcel of

engineering practice is not of direct relevance to the distinction between the two

axiological turns. It is a highly significant fact about the object of study of each axiological turn but does not, as such, undermine the distinction between these two

turns. From the point of view of a descriptive axiological turn the use of thick concepts in engineering practice is an interesting and highly important topic for further

empirical study and from a normative axiological point of view the justification of


Many more items can be added to this list. For instance, most of the “ilities” that play a key role

in software engineering, and outside that field, appear to be thick concepts; see, for instance, http://



P. Kroes and A.W.M. Meijers

the (often implicit) value judgments contained in or implied by the use of thick

engineering concepts calls for a critical analysis.

So, if Williams’ view on thick concepts does affect our distinction between a

descriptive and normative axiological turn, then it must be at the meta-level, the

level of their approaches, not of their object of study. The way we defined both

approaches assumes a fact-value distinction, or more precisely a distinction between

a fact-oriented epistemic and a normative analytic (conceptual) framework. The

approach of a descriptive axiological turn is defined in terms of an analytic framework in which only epistemic values are at work, whereas in the analytic framework

of the normative axiological turn also other (moral, practical, aesthetic) values play

a role. The concepts of the basic values involved in these frameworks, among which

we find the concepts of truth, moral goodness, instrumental goodness and beauty, all

appear to be thin concepts. In order for these concepts to be applied in analyzing

concrete situations they will have to be explicated in more specific concepts that in

turn will have to be operationalized. The concept of truth, for instance, may be

explicated in more specific concepts such as empirical adequacy, explanatory power,

simplicity, coherence etc. and moral goodness in beneficence, doing no harm, pleasure, pain, utility etc. Now the crucial question is whether the application of some

of the more specific concepts in a descriptive axiological turn does or does not, as a

matter of principle or as a matter of fact, involve non-epistemic value judgments. If

they do, then some of these concepts are thick concepts which undermines the idea

that a purely descriptive axiological turn is possible.

We will not pursue the question whether a purely descriptive axiological turn is

feasible any further, except for the remark that we have to be careful not to bring the

fact-value distinction to our analytical frameworks, instead of, in line with Williams’

quote above, “finding it there”.

The issue we have been discussing is, of course, closely related to the longstanding debate about whether science is, or may be in principle, value-free.11 Douglas

(2000), for instance, argues that when it comes to setting standards of statistical

significance in many parts of science the inductive risks involved may have nonepistemic consequences and that therefore those parts of science are not ‘valuefree’. Her arguments also apply to setting standards of (statistical) significance in

engineering practices. At the end of her paper she concludes (p. 578):

Finally, there are cases where the science will likely be useful but the potential consequences of error may be difficult to foresee. This gray area would have to be debated case

by case, but the fact that such a gray area exists does not negate the basic argument: when

non-epistemic consequences of error can be foreseen, non-epistemic values are a necessary

part of scientific reasoning.

If we replace science in this quote by philosophy of technology, and if the philosophy of technology claims to be somehow useful in shaping technology – even if

this “somehow” involves much deeper shades of gray than in the case of science –


This debate goes back at least to Rudner (1953) and the reply by Jeffrey (1956); for a recent

contribution to this debate, see Douglas (2000).


Toward an Axiological Turn in the Philosophy of Technology


then we see no reason why the same should not be the case for the philosophy of

technology. This brings us to the normative axiological turn.


A Normative Axiological Turn: Values and Norms

in Philosophy of Technology

A normative axiological turn implies a departure from the empirical turn. Now there

is no attempt to stay safely within the realm of value-free inquiry (except, of course,

for epistemic values). This turn involves actively taking up normative issues in the

philosophical analysis of the engineering practice. There are at least two options for

a normative axiological turn:

• The reflective position. According to this view it is the task of the philosopher of

technology to analyze normative issues related to technology and engineering

practice and to actively participate in societal discussions by preparing and facilitating debates and decisions about technology;

• The substantive position. In this view the philosopher of technology not only

analyses normative issues but also takes a normative stance him/herself with

regard to the issues at stake and acts accordingly.

It is not our intention to enter here into a discussion of whether these two variants

of a normative axiological turn can always be clearly distinguished and which

option the philosophy of technology, if any, should choose. Instead we will make a

number of observations that are meant to elaborate what we have in mind with a

normative axiological turn.

The first thing to mention here is that the second position clearly makes more

normative commitments than the first. The first position, however, is itself not free

from normative commitments, since there is no value-free analysis and participation

in societal debates: issues are framed in a certain way, there is a choice of aspects

that are taken into account, some values (debates) are considered to be more important than others, etc. But these normative commitments play a role in the background and are about the way the normative issues at stake should be analyzed,

debated and decided upon.

The second thing to observe is that taking a normative axiological turn does not

necessarily imply taking a moral turn. The normative axiological turn may pertain

to all kinds of values and norms: epistemic, moral, practical, aesthetic etc. As we

noted in our empirical turn paper, there has been a long normative tradition in the

philosophy of science of a non-moral nature. In this tradition various schools within

the philosophy of science have criticized the epistemic values and methodological

norms applied in actual science on the basis of the epistemic values and methodological norms accepted by those schools. One can easily imagine something analogous to happen in the philosophy of technology. A philosopher of technology may

criticize on the basis of her/his epistemic values and methodological norms a


P. Kroes and A.W.M. Meijers

particular engineering practice for believing that an engineering theory is sufficiently supported. In that case there is a clash between two different sets of epistemic values and norms. Similarly, starting from some set of values and norms for

practical, instrumental action a philosopher of technology may criticize the use of a

particular theory as a reliable guide for action. And, of course, by taking a moral

stance a philosopher of technology may criticize moral principles of and decisions

taken by engineers. By taking a moral stance not only principles and decisions that

are moral in nature, but also morally relevant claims and decisions may be criticized. Within a particular context of action, believing a theory to be sufficiently

supported for accepting that theory as a reliable basis for action may be highly morally relevant. In other words, taking a moral normative stance may involve or may

make it necessary to take also an epistemic and practical normative stance. This is

what makes taking a normative stance such a complicated affair. As we noted earlier

the “standards of excellence” of engineering practice cannot be dissected into independent sets of values and norms that correspond to the neatly defined values and

norms of standard philosophical domains.

Third, if the philosophy of technology is taken to be itself a normative practice

that is based on a substantive moral stance, then the question about its basic (morally relevant) ends and values arises. What are these values? For medical practices

they are about curing diseases and improving the health of patients. For engineering

practices these basic values are contained in their codes of conduct which usually

contain phrases like “Engineers shall hold paramount the safety, health and welfare

of the public in the performance of their professional duties”.12 In analogy to the

ABET, that defines engineering as a profession for “the benefit of mankind”, the

philosophy of technology may conceive of itself as a discipline for the benefit of

mankind. But, of course, this is not of much help; it does not tell us anything about

how this basic value shapes the discipline.

If the core activity of a normative philosophy of technology is critical reflection

on technology and taking a reflective or substantive normative stance on the issues

is at stake, then the epistemic, practical, moral etc. values and norms underlying

critical reflection itself have to be critically questioned at the meta-level. The result

does not necessarily have to be a set of (morally relevant) substantive values and

norms that have to be adopted by the discipline; these values and norms may also be

of a procedural nature and involve a procedural role for philosophers of


A fourth and final point, related to the foregoing one, is that a normative axiological turn in the philosophy of technology will inevitably run up against what is

known as the is-ought problem or as Hume’s guillotine. This problem is the more

pressing the more normative philosophical commitments are involved. In that sense

it may affect the reflective position less than the substantive position. So far, we


See, for instance, the code of conduct of the NSPE: http://www.nspe.org/resources/ethics/codeethics, and of the ASME: https://community.asme.org/colorado_section/w/wiki/8080.code-ofethics.aspx.


Here we touch upon the ethics of philosophy as a profession; see (Hansson 2015).

Tài liệu bạn tìm kiếm đã sẵn sàng tải về

4 Is a Purely Descriptive Axiological Turn Feasible?

Tải bản đầy đủ ngay(0 tr)