Modeling artifact use

34
0 Modeling artefact use trough philosophy of action An analysis at the intersection of philosophy of technology and philosophy of cognition German Dutch workshop Philosophy of Technology Darmstadt June 2014 Ir. Hans H.M.Tromp, MA Radboud Universiteit Nijmegen

Transcript of Modeling artifact use

0

Modeling artefact use

trough philosophy of action

An analysis at the intersection of philosophy of

technology and philosophy of cognition

German – Dutch workshop Philosophy of Technology

Darmstadt June 2014

Ir. Hans H.M.Tromp, MA

Radboud Universiteit Nijmegen

1

Action with artifacts v2d

Contents 1. Introduction ......................................................................................................................................... 1

2. Philosophy of action ............................................................................................................................ 2

2.1. Introduction ................................................................................................................................. 2

2.2. Actions, Intentions and Reasons .................................................................................................. 3

2.3. Causal explanation or understanding of actions .......................................................................... 7

2.4. Conclusions ................................................................................................................................ 10

3. Views on Artifacts ............................................................................................................................. 11

3.1. Introduction ................................................................................................................................ 11

3.2. Function theories and intentional approaches of artifacts ........................................................ 11

3.3. The Interaction views ................................................................................................................. 15

3.4. Philosophy of mind ..................................................................................................................... 18

3.5. Conclusions ................................................................................................................................ 21

4. Modeling Action with artifacts .......................................................................................................... 22

5. Conclusion ......................................................................................................................................... 23

Appendix 2.1. Explaining versus understanding............................................................................ 24

Appendix 2. Feedback loops in action theory ............................................................................... 26

References ......................................................................................................................................... 29

References for section 2 ................................................................................................................ 29

References for section 3 ................................................................................................................ 31

1. Introduction

The purpose of this paper is to verify the application of the modelling concept of artifacts, as in

developed for my thesis ‘On artifact design, through philosophy of cognition and action’ in relation to

action theory. For that purpose, in section 2 a concise modeling of actions will be derived from action

theories as available from philosophy of action. In section 3 the modeling of artifacts as presented

earlier is summarized. (Tromp 2013). Then in section 4 the combination of the two models will be

discussed. Section 5 concludes the results.

2

2. Philosophy of action

2.1. Introduction The purpose of this section is to establish the applicable definitions from available philosophical

references on actions and causal concepts of action in order to analyze and model actions as an input

for the relation with artifact use.

Philosophy of action, also indicated as action theory, covers a range of increasing philosophical

activities stimulated by recent results in neuroscience and by related discussions on free will and

moral responsibility. The philosophical views and approaches to these issues differ considerably. In

this exploratory study the focus will first be on what can be indicated as the operational cognitive

and action theoretical aspects. 1 Action theory started to be perceived as a separate subject halfway

through the previous century by Wittgenstein's introduction of a very basic question in action theory:

I raise my arm, my arm goes up. And the problem arises: what is left over if I subtract the fact

that my arm goes up from the fact that I raise my arm (Are the kinaesthetic sensations my

willing?).2 (Wittgenstein 1945: 621)

Among the many philosophers who have worked on action theory, Elisabeth Anscombe, a close

friend of Wittgenstein, and Donald Davidson played a dominant role. Donald Davidson can be

considered as the initiator of the search for the causality of actions. His view is discussed below in 2.2.

Elisabeth Anscombe's Intention is considered to have firmly established the philosophy of action as a

distinctive field of inquiry as among others discussed in Ford, Hornsby and Stoutland’s Ascombe’s

Intentions in Context, Essays on Anscombe's Intentions (Stoutland 2011) Anscombe's views will be

considered in 2.3 under Intentions.

To structure the complexity of action theory, Alfred R. Mele's structure from his introduction as

editor of The Philosophy of Action will be used here. The central questions in philosophy of action,

according to Mele are: What are actions? and How are actions to be explained? (Mele 1997)

Philosophy of action is about actions, causes, reasons, intentions and the explanations, and

understanding of their relations. Various action theories appear to apply different meanings to terms

depending on the objectives of their investigations they are a part of. In his introduction of The

Philosophy of action Alfred A. Mele identified 7 different (causalist and non-causalist) views on the

relation between these terms. (Mele 1997, 4–16) However, the purpose of this section is not to give

or summarize an overview of existing theories, but to find the basis for the analysis of artifact use in

actions with references to elements of available theories. Therefore, first the choice of meaning of

these applicable terms as applied in this paper in relation to my position will be defined.

For daily use it is not necessary to distinguish very strictly between explanation and understanding,

but for this kind of philosophical discussion it is useful, although not practiced in general, to

distinguish the meaning of these two terms. The use of these terms is related to the reductionist

1 The relation to notions of free will and moral responsibility will be outside the scope of this paper.

2 Brackets in original text.

3

versus non-reductionist positions in the philosophy of mind debates. Jennifer Hornsby states that

human actions are not (fully) accessible from an impersonal or neuroscientific point of view, and

cannot be reduced to a physiological level. (Mele id: 6) This is in line with the position I take in this

paper as mentioned above, which can be indicated as a two-level one: conceptual at the higher level

of understanding actions, and reductionist when explaining actions at the lower level.

For more details and justification of this position and for the use of the terms explanation and

understanding see appendix 1.

2.2. Actions, Intentions and Reasons The three dominant terms Actions, Intentions and Reasons in action theories are interconnected in

various ways with a strong - or a weaker forms of causation. Within the applied concepts intentions

and reasons, the terms beliefs and desires are frequently used and therefore need some attention as

well. As it is not possible to give a good overview and analysis, this section gives only a short

impression of the views of some main players in this field of philosophy on the causal relations.

What are actions

Quite some authors refer to Donald Davidson's definition in Agency (1971) that actions are bodily

activities. This definition would be applicable for the lower level users aspects, but is too narrow if

one would include the deliberation processes, usually thought of as “practical reasoning” that can

precede the actual action. (Davidson 1987:106)

Action-individuation can be done by taking different views (Mele 1997):

Fine-grained view: considers each action with different characteristics separately. For

example turning the key of the car and starting the car are different actions.

Coarse-grained view: considers the two actions of the example above as being one action

under two descriptions.

Component view: regards starting the car as an action with various components, as moving

the hand towards the key, etc.

These views have been a subject of discussion since the end of the 1970's, but Mele agrees with Carl

Ginet's statement that this issue is not one on which much else depends. (Mele 1997: 2) In this paper

each of the three views will be used when applicable, for although there is not much difference from

a conceptual point of view as Ginet and Mele state, they are at times quite different from a cognition

point of view.

Actions are explained mainly in two different ways, namely by looking at reasons or at causes. Some

theories require causal explanations as law similar to physical laws. Then reason-explanations are not

causal explanations. (Mele 1997, 4) This seems to contrast with theories on acting for reasons.

(Rober Audi in (Audi 1997) At least part of this problem is in the meaning of explanation versus

understanding and of cause, as indicated above.

4

In his Problems in the Explanation of Action Donald Davidson picked up Wittgenstein's basic question

and stated that it is the agent that raises his arm, but then that [t]he cause can't be identical with the

effect. (Davidson 1987: 102) There should be something that makes the difference between what

happened to the agent and what the agent does. Davidson explains this by an example of raising a

paralyzed left arm by pulling on a rope via a pulley with ones right arm. If he were to do that, the left

arm would be raised by doing something else.

He then concludes:

Perhaps it will now seem that after all something is added to an event, say my arm going up,

is an action, then there must also be an intention. But in my view, the intention is not part of

the action, but a cause of it.

[So] nothing is added to my arm going up if that event is caused by an intention.

At one time (about twenty-five years ago, when I wrote ``Actions, Reasons and Causes'') I

thought there were no such states as intending; there were just intentional actions. This was,

I now believe, an error.

Intentions are also required to explain how complex actions are monitored and controlled.

(Davidson 1987: 106)

This last sentence is an important remark, as it suggests a control loop as mentioned below, but

Davidson did not work it out.

In action theory it is quite common to consider beliefs and desires as the main elements of

intentionality: intentionality is the desire to accomplish a goal on the basis of a belief. Belief in the

context of action theory means knowledge and/or expectations about the world, to reach that goal.

Although it can be questioned whether this concept is always valid it seems to be applicable in the

cases we will consider in the next section.3

Davidson stated that the emergence of an intention requires two transformations with regards to the

belief and desire. First, the transformation that brings belief and desire together. This is what is

usually thought of as practical reasoning. Second, the transformation in which the perceived end-

value is balanced against the value of the means. (Davidson1987: 106)

Fred Dretske picked up the point of the mental causation as mentioned, but not fully worked out, by

Davidson. In Explaining Behavior Dretske analyses the processes in terms of beliefs and desires as the

elements that play a role. (Dretske 1988) However, his diagram with its explanation deviates from

the conventional way of analyzing these concepts and does not mention the essential key of the

concept, being the feedback loop. (see appendix 2)

3 Animals and even insects show behavior that we would interpret as intentional, which is most probably not

based on separate notions of belief and desire in their neural systems, but predefined by DNA-driven sets of causal processes. Therefore we cannot exclude similar mechanisms in human behavior.

5

Davidson rejects the physiological oriented approach to relate the physical and the mental domain of

philosophers as Dretske. (Davidson1987: 115) However, as considered below, at different levels both

contribute to the understanding of actions.

Intentions

The term intention, frequently used in philosophy of action, needs some attention.4 As cited above,

Davidson finally concluded that intention causes the action and he indicates intention as the

combination of belief and desire. Much earlier, on the cover of the Harvard University Press edition

of Elizabeth Anscombe's Intention, he wrote [this] is the most important treatment of action since

Aristotle. (Stoutland 2011: 1) However, in his final essay Problems in the Explanation of Action

Davidson did not make any reference to Anscombe. (Davidson 1987)

Anscombe gives us an account of what distinguishes intentional actions from other events, that they

are actions to which a certain sense of the question why? applies. (Stoutland 2011: 26)

In her contribution to the Essays on Anscombe's Intention Jennifer Hornsby concludes: I think that

neither Anscombe nor Davidson was fully aware of the distance that there was between them.

(Hornsby 2011, 126) Hornsby rejects what she considers the too much event-oriented position of

Davidson in favor of the more reasons why position of Anscombe. She concludes that we must give

agents the central place in a story, and allow that agents have powers to affect things, and that what

they bring about intentionally is shaped by their reasoning. (Hornsby 2011, 125)

To illustrate her view on Davidson's position Hornsby cites Davidson speaking of actions as

"sandwiched between cause and effect". Davidson illustrates that with:

If we know that someone intentionally crushed a snail, we know that some action of his was

caused by a desire to crush a snail, and a belief that by performing the action he would

promote the crushing of the snail. (Davidson1987: 108)

It is striking that Davidson does not mention a reason to crush the snail. This is typical of Davidson's

examples, usually kept to low level bodily actions.

I fully agree with Hornsby that explanation of actions with the focus on bodily movements should be

considered in the wider context of an agent's activity. 5 To elaborate on Davidson’s example above: a

gardener will crush a snail in his garden to defend his plants in the larger context of maintaining his

garden, while he will leave the snail untouched when walking a nature trail.

As mentioned above, Davidson took a first step in the direction of explaining the how of an action,

that complements the why approach of philosophers like Elizabeth Anscombe.

Acting for Reasons

Not all authors give a clear definition of the difference between intention and reason. Wayne A.

Davis refers to Davidson (1963) with the definition of action for a reason (without any reference to

4 The term intentional stance as used by Daniel Dennett in relation to artifacts will be considered in section 3.

5 Jennifer Hornsby proposed the expression activities during the seminar Acting with Artifacts at the Technical

University Eindhoven NL, November 2012.

6

the notion of intention): The causal theory of reasons maintains that an agent does something for a

reason only if the action results from a certain belief and desire. And: agents act for reasons only

when their actions result from practical reasoning. Agents’ own rationalization afterwards are

explicitly not included. (Davis 2010: 34-35) The causal theory of reason also covers the cases where

reasons for action are false.

Robert Audi states: Clearly acting for a reason is closely related to acting intentionally. He discusses

the question whether what he calls intrinsically motivated actions are actions for reason, with the

example of someone who goes for a short walk. This person may reason to be in need some fresh air,

or might just feels like strolling. Audi argues the question is important because, if intrinsically

motivated actions are actions for reasons, then intentional action is equivalent to actions for reason.

He concludes that action for a reason can be distinguished from reasoned actions, being the result of

practical reasoning. (Audi 1997: 103–104)

An example of an action for reson could be: Seeing someone standing in a train raising an arm (and

gripping a hold) in case of a speed change is directly understood to have the reason to avoid falling,

but here reason will not be considered as a cause. As it can also be explained as a reflex, being a set

of physiological processes learned / developed as an effective way to maintain balance.

Davis also notes that a virtue of the causal theory of reasons is that it provides a simple account of

why we can explain a person's action by citing the reasons, while citing the reasons he had for doing

them may not be sufficient to explain them. (id. :35)6

In practice there can be a complex mixture. S might be participating in a not too strict formal meeting,

after about an hour she might get the bodily feeling to stretch the legs, to need some fresh air, while

the same time noticing the discussion moving in a way she does not like, as a result she sees a good

opportunity to propose a break.7

Beliefs and Desires

The terms belief and desire and their interrelation as elements used in the above mentioned

approaches are often not worked out in detail, so a short summary of the results derived from above

as used in this paper.

Beliefs

In Practical Reasoning Gilbert Harman defines intentions as a result of practical reasoning are

psychological states that are not to be analysed away in terms of reasons, beliefs, desires and

behavior. (Harman 1997, 150) But then he applies the term belief in two ways. First he relates belief

to intentions as the conviction that a considered action will be performed. Secondly he note that the

result of certain practical reasoning activities resulted in the belief that the intention will be realized.

As practical reasoning is reasoning about what to intend and as theoretical reasoning is about what

to believe, he states that this belief then belongs to the domain of theoretical reasoning. (Harman

1997: 152) This leads to a complex interwoven structure of theoretical- and practical reasoning in

6 Note that Davis takes a third person interpretation view here.

7 This example is build on a short one from Robert Audi.(Audi 1997: 103)

7

particular in multi level activities. Although this might be of interest in certain philosophical domains.

In the context of this cogni

In this paper I remain closer to Davidson. As mentioned above, Davidson assumes a transformation

that brings desires and beliefs together in a practical reasoning process. Practical reasoning includes

all reasoning of means to end, including the evaluation of alternatives and value balancing.

Theoretical reasoning will be used with the common meaning as reasoning about the knowledge of

the world and will result in the belief component of reasons and intentions. Here I follow Harman:

the intention is the integrated structure of belief and desire, but with the additional remark that this

structure is in control of the performance of the action in progress and that deviations of the result

are connected to higher, corrective levels of action as indicated in the example below and in the

example of the Vestibulo Oculair Reflex as described in appendix 2.

Desires

Desires as causal elements of an action in many cases belong to a hierarchical chain of desires. For

example the desire to read a particular book can be related to the desire to prepare a paper, to pass

a certain exam, which follows from the wish to reach a certain status. But in parallel the reading can

be based on the desire to acquire the knowledge offered by the content of the book. Harman

distinguishes intrinsic desires from extrinsic desires. (Harman 1997, 165–167) Intrinsic desires are

unanalysable attitudes in their own right (id.:165).8 Extrinsic desires are explicit elements in practical

reasoning. In the example of reading a book the desire to reach a status and the desire to acquire

knowledge are good examples of intrinsic desires, while the others are elements in practical

reasoning processes and therefore related to different types of cognitive processes.

Both intentions and reasons are assumed to be based on beliefs and desires as causal elements. In

practical reasoning each of these two components play a separate role as well as being interrelated

in the evaluation of alternative actions or ways to perform such actions. Intentions can be considered

as the consolidated or routine result of pragmatic reasoning; desire and beliefs are then integrated

elements in the process of action.

2.3. Causal explanation or understanding of actions

As mentioned above in this paper explaining will be restricted to the use in basic functional processes,

preferably related with physiological processes, but with an extended use of the term cause (see

below). Understanding will be used for 'explanations' by reason, with the exception of the first

person use citations.

Wayne Davis claims that the basic principle common to all causal theories of action is that:

[t]he agent performs an action only if an appropriate internal state of the agent causes a

particular result in a certain way. (Davis 2010: 32)

8 Unanalysable here should be understood in terms of analytical reasoning.

8

Wayne A. Davis observes that there is a distinct difference between agent causation and event or

state causation. This results in a variety of views. To summarize these views very briefly: agent

causation makes use of the concept of agent intention, in which intention usually is formed by the

combination of beliefs and desires. Other views put emphasis on the agent as a complete entity.

(Davis 2010: 32-34) With the event- or state causation approach the agent intentions are not the

main interest. Event causation seems to be typically applicable in situations where the agent is

subject to an unexpected external situation and reacts to that situation in a reflex or pre-considered

action. In the context of this paper, state causation is the more applicable approach. However,

internal states will not be considered as static situations, but as process configurations as described

in the examples of appendix 2.

Modeling action causation

Actions are processes, and as such they are not to be considered as ontological entities but as mental

constructs.

From the above-mentioned theories we can define the following classification of actions into three

levels:

Practical reasoning

The highest level are the deliberate considerations usually classified in the domain of

Practical Reasoning. Although Ellijah Millgram defines practical reasoning is figuring out

what to do. (Millgram 2012) However, in the context of this paper also the how to do

considerations are included in this cluster of mental activities, but the with free will related

normative considerations are not in the scope of the analyses made here. The choices on

between desires and the possible ways, based on beliefs, to realise these result in intentions.

Intentional level

Intentions can be considered as a combinations of a desire and a belief that controls and

monitors the detailed actions that belong to the activity that aims to realize the desire with

the means as assumed in the belief.

Sensory-Motor level

The direct bodily execution of the actions are controlled can be quite complex but basically

well defined neurological feedback loop processes like the Vestibulo Oculair Reflex example

in appendix 2.2

This three level clustering is very similar to the concept with distal intentions, proximal intentions and

motor intentions defined by Elisabeth Pacherie. (Pacherie 2008) But, the naming remains more close

to the usual indications in philosophy of action.

The diagrams below visualize the interrelations of this modelling.

9

fig.2.1 three level action model

fig. 2.2. multilevel feedback model

Example:

The approach followed in this paper can be explained by the following example:

1. I want to read a paper while , for instance to prepare a presentation for this seminar.

2. At the lowest level, the Vestibulo Oculair Reflex compensates for small movements of the book.

3. With light bumping I will keep my arms free from the armrest due to a subconscious reaction.

4. With slightly heavier bumping I will keep the book closer to my eyes and probably be aware of doing this.

5. When the bumping becomes stronger I will stop reading and make an estimation on the probability of

improving conditions.

6. Next I might decide to put the book aside and start reading when conditions improve. This would be a

deliberate decision.

10

2.4. Conclusions

Philosophy of action offers various approaches on actions, intentions, reasons, desires and beliefs

with various views on their causal relations. For the purpose of the analysis in this paper the

following conclusions can be drawn:

C1. Action should be understood as an element in an hierarchical chain of purpose-directed activity.

Lower levels of the chain are controlled and monitored by higher levels.

C2. Low level bodily movements can be explained as being controlled by hierarchical neural

feedback-loop configurations.

C3. At the higher level, actions can be understood as Acting for reasons, the result of practical

reasoning in terms of beliefs and desires.

C4 It seems reasonable to assume the same type of multilevel neurological processes are applicable

at the levels between the reasons as output of practical reasoning and the bodily actions as

intention-controlled feedback systems. However, due to complexity of the neuron networks only a

functional understanding is feasible.

C5 Intentions in this paper are to be considered as interrelated consolidations of desires and beliefs,

but each under control by higher levels.

C6 There is also the general inclination of human beings to interpret phenomena as caused for a

reason. This inclination will be considered in the next section.

C7 The highest end remains in the vagueness of free will motives out of the scope of this paper.

C8 Something still is missing from a phenomenological / pragmatic point of view actions, and

certainly actions with artifacts as analyzed here, are not just a performance of agents. Actions, in

particular actions with artifacts, are actions in and with the world. These aspects will be worked out

in the next sections.

C9 Although philosophers use the example of a thermostat, (Davidson, Dretske, Dennett) a lack of

understanding can be observed about the key concept of negative-feedback-loop-control-systems.9

9 This is caused by the fact that this concept is not included in the standard curriculum of physics, as in physics (in contrast

to biology) these systems play no role. With the typical example of the application of thermal extension, the bimetal-based thermostat hides the concept of a negative-feedback-loop by the integration of measurement, reference levels and switch function.

11

3. Views on Artifacts10

3.1. Introduction

The purpose of this section is to summarize those views that are relevant for the definition of a

reference model in relation to actions. A number of views on artifacts focus explicitly or implicitly on

the ontology of artifacts, often in contrast or analogy with the ontology of natural things, and

although not a metaphysical study, some ontological concepts are worth considering as input for the

analyses to be made.

Recent philosophy offers a range of views on artifacts. These views can be clustered into three main

categories, namely:

1. The functional and intentional theories; these theories belong mainly to the domain of an

analytical approach.

2. The interaction views; the effect of the artifacts on the user and the related interaction.

These views belong more to pragmatism and phenomenologically oriented philosophy.11

3. The Theory of Mind approaches; this third cluster highlights the philosophy of cognition

approach.

It appears that these categories of views are almost complementary. As these categories have been

developed in different contexts, philosophical debates on artifacts are mainly restricted within these

categories. First, the views on artifacts clustered in these three categories are summarized. Then, the

views are used to define a basic development model.

3.2. Function theories and intentional approaches of artifacts Some authors consider functional and intentional views as opposite approaches to characterize

artifacts, others seek for combined approaches of the notions function and intention. In this section a

number of functional, intentional and combined views are reviewed in order to identify the

applicable input for the analyses of this paper with the focus on actions with artifacts.

Functional theories

A good number of views from different disciplines and approaches support the notion of function as

a main characteristic of artifacts. In Practical Reasoning and Engineering Jesse Hughes states that

artifacts are typically categorized according to function, that is according to the ends for which they

are intended as means. (Hughes 2009: 376) He illustrates this with an example of fire extinguishers. If

10

This text is a subset of my presentation at the SPT2013 conference in Lisbon.(Tromp 2013) 11

This observation of the relation between these two clusters and the respective philosophical disciplines is also made by Auke Pols in his Acting with Artifacts (Pols 2011: 78).

12

you want a fire extinguished, you can use a fire extinguisher. He then adds in a footnote: knowing the

artifact’s function also suggests when it should not be used. 12

This notion function is also the key in the description of design details and the identification of parts

in drawings, as used by Peter Kroes' analysis of the naming of artifacts with the example of an

exploded view of a car carburettor. Part identifications such as: accelerator pump cover, pump

discharge nozzle, throttle lever etc. indicate the functions of the parts in the total. (Kroes 2006, 138)

Judit Futó, Téglas, Csibra and Gergely show that structuring in relation to function is not only related

to language but seems to be a basic cognitive mechanism that can already be observed with

preverbal infants: function demonstration can induce kind assignment and object individuation in 10-

month-old infants even in the absence of linguistic labeling. (Futó et al. 2010: 8)13

According to Beth Preston, most of the discussion on function theory is in the domain of philosophy

of biology, but artifact function received comparatively little attention: 14 there is no breeding ground

for theories of artifact function. She states that a full-fledged function theory of artifacts ought also

to account for a variety of considerations (Preston 2009), applicable in this context15:

Multifunctionality: for efficiency reasons an artifact might be designed to provide more than one

function. The Swiss army knife, the combination-pliers and the multimeter (voltmeter with some

added circuitry to measure also electronic current and resistance) are typical examples.

As Preston remarks, recycling can be considered as a special case of multifunctionality.

Concurrent functionality: artifacts might fulfill functions in different domains at the same time.

For example:

o An uniform has in addition to the function of clothing a communication function to

indicate the belonging to a certain group, with a certain rank.

o A building often is more than a functional shelter, it will be an investment and often

indicates the status of the owner.

o Status and symbolic functions: examples given by Searle are: traffic lights, money, a

wedding ring. The status function can often be combined with another function.

Accidental, user invented functions: functions the artifact was not intended for, like an umbrella

designed to protect against the rain, which can also be used against sunlight or as a weapon.

Closer to daily use is the spoon used to open a cocoa tin. (Preston 2009, 214–215) 16

12

This remark will be used in the analysis in section 4 13

Infants and artifacts will be one of the subjects of chapter 7. 14

Although there is some similarity with the use of function in this paper, the discussion on function in biology does not

contribute to the analyses here. The focus in philosophy of biology in the discussion of function is on the teleological aspect of evolution. A good survey of the various aspects can be found in Collin Allen’s Teleological Notions in Biology (Allen 2003). 15

This listing is based on (Preston 2009) but slightly differently configured and with some examples added. The main

difference is that I split Preston’s multifunctionality in multifunctionality and concurrent functionality, as specified here. 16

I had in mind the screwdriver used for other functions like opening a paint can.

13

Peter-Paul Verbeek articulates this last one more explicitly or even extremely as the ultimate

users view: the artifacts do not have an essence, they should not be considered outside the users

context, they are what they become in usage (Verbeek 2000, 134). His view within the tradition

of postphenomenology will be discussed below.

The above-made references support the functional approach. However, specifically in the

metaphysical position a set of problems arises. The functional approach inevitably leads to the issues

of defect, broken down or malfunctioning artifacts.17 Several ontologically oriented authors require

of functional theories of artifacts that the question is answered whether a defect artifact is to be

considered as an artifact. For a pragmatism oriented analysis this does not seem to be a problem. A

defect car will be indicated as a defect car, or when it is serious it will be a broken down car, and

when it has been damaged at the outside it will be a damaged car; finally when it is hopeless it will

not be a car anymore but indicated as a wreck.

Intentional approach

In her The Metaphysics of Everyday Life Lynne Rudder Baker seeks the solution of malfunctionality in

the notion of intention (Baker 2007). She defines: Most broadly artifacts include everything that is

produced intentionally. (id. :49) This view focuses on what the artifact is supposed to do and

implicitly introduces normativity. So Baker remarks: nowhere (in philosophy) is normativity more

glaring than in the behavior of artifacts.18 This intentional approach avoids the functional view's

problem of the malfunction, because there is no intention without the possibility of its being

frustrated, no function without the possibility of malfunction (Baker 2007: 51). The definition of the

proper function of an artifact is the intended function, even if the artifact never performs its proper

function, the ontological problem of malfunction is avoided. (id.:52)

Still, a strict intentional view comes with some other problems. The possibility that an artifact can be

used for other than the intended function, the accidental function as mentioned above cannot be

covered by an intentional theory. The reference to intended functions includes the design

perspective or the supposed design perspective. This makes it very dependent on the knowledge of

the intention of the creator and then it is not uncommon to connect the two approaches by referring

to proper function as the intended function. This meets the ontological oriented requirement that

artifact existence can only be understood by the assumption that the originator had an intention as

Baker stated. A user-invented (accidental) function then is to be understood as a new invention and

can be at the start of a chain of new developments.

Combined functional and intentional theories

Peter Kroes and Anthonie Meijers combine the two notions function and intention, in their definition

of a dual nature concept of artifacts as a kind of paradigm for the introduction of a major

philosophical research program19:

17

See for some more detail about malfunctioning Misrepresenting and Malfunctioning. (Neander 1993) 18

The notion behavior will play a major role in the conceptual analysis in section 5. 19

This research program, supported by the Netherlands Organisation of Scientific Research (NWO), was a joint activity of several universities between 2001 and 2006, initiated by the Delft University of Technology - see http://www.dualnature.tudelft.nl

14

In so far as technical artifacts are physical structures they fit into the physical conception of

the world; in so far as they have intentionality-related functions, they fit into the intentional

conception. Both conceptions are necessary for characterizing technical artifacts.

[t]he best way to conceive the functions of artifacts, is to regard the notion of function as a

bridge concept that relates the physical and intentional domain. (Kroes and Meijers 2006, 2)

But they acknowledge that there is no generally accepted theory of function available in literature.20

In this research program of the dual nature of artifacts, Wybo Houkes' and Pieter E. Vermaas'

developed their Intention, Causal, Evolutionary (ICE) function theory as a function theory for

technical artifacts based on an action-theoretical analysis 21. (Houkes and Vermaas 2010). They

evaluated three existing theories, with their own shortcomings, (id. :2) and aim to combine the

Intentional theory with the Causal-Role function theory of Robert Cummin and the Evolutionist

function theory, using the useful elements of each, and by doing so avoid the shortcomings of each of

these theories as mentioned above.

They specify (id.: 18):

o Plans are deliberate efforts to change the world, as such, they are based on beliefs about

the world, ourselves, and the effect of actions.

o Plans as complex, mental items consist of considered actions, not of actual actions.

o If artifacts are involved the plan is called a use plan.

o The mental process of planning is constructing or designing a use plan. This process

results in a more or less durable mental state, similar to a belief or an intention.

They add that plan in daily use already includes a good amount of (practical) rationality. This includes

the following context-related elements: effectiveness, goal consistency, means-end consistency and

belief consistency.(id.: 39-41).22 These elements of rationality can then be used to evaluate artifact

use and design. The Intentional theory element is still present as use plan. Use plan and intentional

theory describe artifacts primarily in a teleological context. Agents consider artifacts as a means to

realize goals. But use plan is more restrictive than the intentional theory as it affirms the belief that

the artifact will contribute to the required goals. This is called effectiveness belief. (id.: 79)

Peter Kroes elaborates in his Technical Artefacts: Creations of Mind and Matter on the dual concept

of artifacts and discusses the ICE function theory as mentioned above. (Kroes 2012) In his chapter 2

on technical Artifacts he seeks a conceptual framework to relate the physical and social/intentional23

20 This is probably due to the attempts to combine the concept of function in the domain of biology with that in

the domain of technology.

21 Intention is put in front as the most important element of the theory.

22 In the domain of action theory, belief is usually combined with desire; in most cases and certainly in this context it has

the meaning of knowledge. 23

In this paper I will follow the plural concept of functionality as defined above. I therefore include Searle’s social aspect

and will stick here to intentionality only, social aspects will be considered separately.

15

conceptualizations of part of the world (id.:19). The dual structure makes the technical artifact a

hybrid kind of object (id.: 40). The intentional, means-end structure has the structure:

x is for ϕ-ing, to be an artifact of the ϕ-kind it is considered to be a ϕ-er (id.: 37).

This structure is extended in a normative way by evaluative statements as:

x is good for ϕ-ing / the object is good for driving screws, and prescriptive statements as norms of

behavior. (id.: 89) 24

Kroes adds a relation to the context of human actions to the dual nature of artifacts with reference

to Herbert A. Simon and visualizes the relations in the diagram below:

fig. 3.1

fig. 3.2

fig. 3.3

These artifact conceptualizations as visualized in the diagrams will be used as input for the

conceptual analysis of subsection 4.

Conclusions

C2.1.1 Both function and intention are applicable approaches to understand artifacts, be it from two

different perspectives.

C2.1.2 Functions are recognized as cognitive attributes to artifacts.

C2.1.3 Artifacts in general should be considered with a plural notion of functions, as listed above.

C2.1.4 Intention as a notion in relation to artifacts is applicable when the purpose of the designer of

the artifact is to support a human activity.

C2.1.5 The notion of use-plan places the artifact in the context of an activity.

C2.1.6 The dual nature of artifact concepts should be extended with an environmental / context of

human action aspect.

3.3. The Interaction views A number of views on artifacts share the perspective that artifacts have to be understood in their

interactive relation with the subject. In this cluster three sub-clusters can be identified.

Pragmatism and Phenomenology

Phenomenology and pragmatism as two independent major philosophical developments of the first

half of the last century share not only their historical context as a reaction against Neo-Kantianism

and their revival at the end of that same century, but what is more important to this analysis, they

24

The use of the term behavior will be considered in section 4.

16

share their anti-dualistic conceptual framework with the interrelation of subjects with the objects in

the world as common main paradigm.

In Sein und Zeit Heidegger identified functional artifacts as being available (vorhanden) in an action

oriented meaning in the environment (Umwelt). (Heidegger 1927: 1.3.15/69) For this he found

support of a typical characteristic of the German language with the combinations of words with zeug

such as: Schreibzeug, Nähzeug, Fahrzeug and Messzeug.25 This expresses the action relation with the

object. Zeug puts objects into categories of action such as sewing gear or hunting gear. Constitutive

for the equipment is its in-order-to (um-zu). Heidegger emphasized that the interrelation not only

consists of knowledge but also of experience using the example of the hammer. Using the hammer

discloses the specific handiness of the hammer (id.:69). This interaction accounts for the diversity of

hammers.

Don Ihde can be considered as one of the first North American philosophers who developed a main

interest in philosophy of technology. (Ihde 1983), (Ihde 1991) Although he coined his work as post-

phenomenology, his approach is equally based in the tradition of pragmatism. As Carl Mitcham states:

Unlike many phenomenologists, Ihde has been in constant dialog with pragmatism. His creation of a

post phenomenology is in effect a pragmatic phenomenology. (Mitcham 2006: 31) In his post-

phenomenological work Ihde addresses the fact that artifacts change our relation with the world.

So far the views considered in this section described the role of artifacts in the realization of wanted,

purposive actions of agents. While Ihde focused on the effect of artifacts on our perception, Bruno

Latour and Peter Paul Verbeek give specific attention to the reverse processes. Artifacts influence

and might even cause the actions and the behaviour of people.

In his main work What Things Do Verbeek states that artifacts invite or inhibit certain actions or

behavior, with the example of a speed bump that inhibits fast driving and invites slower speeds.

These are considered as mediated actions.

The concurrent function of the uniform mentioned in section 2 changes the behavior of the person in

an authority sensitive situation, and it would also prevent certain behavior. For example: a

pedestrian to a police officer, both waiting for a red pedestrian crossing light, with no traffic in sight:

`I would not be waiting if you weren't here'. The officer responds: 'same for me'.26

Although Verbeek does not use the term, it seems that his position is more or less compatible with

the notion of affordance, as considered below.

Affordance and functional behavior structure

The affordance concept is based on the general model of interaction with “behaviors” of objects. The

term affordance is introduced as a socio-economic frame by J. Gibson (1979), and the affordance

model places an actor in an environment.

The affordance concept or interaction view is easily recognized. For example, put some paperclips on

a table in a small meeting room and few people will resist the temptation to play with these

25

Eng.: writing materials, sewing kit, vehicle and measuring tool. 26

As read in the column 'ik' in the Dutch newspaper NRC.

17

paperclips lying in front of them. The concept of affordance is specific to artifacts. On a hot day a tree

affords a welcome shade and in a way one could see this as an invitation to take advantage of the

shade. This effect of affordance can be felt stronger in the case of artifacts, because the recognition

of the specific function might cause a stronger invitation for use than natural objects.

Joel Norman distinguishes two visual systems that provide an affordance process and relates these to

neurological structures:

1. The dorsal system deals mainly with the utilization of visual information guiding one’s

behaviour in one’s environment.

2. The ventral system deals mainly with the utilization of visual information for “knowing” one’s

environment.

An artifact as an element in its environment can lead to various acts: a chair can be used to reach

high shelves, or it can afford ventilation by propping open a door. The effects of these interactions

has been dubbed the behavior of an artefact. This is based on the concept that a physical object has

an infinity of behaviors as it interacts with an uncountable set of other objects in the world. A chair

changes local heat transfer, causes a shadow; placed in the right solid position it transfers the weight

of a human to the ground, and at the same time supports the human body to an applicable level, etc.

(Wood 2009 )

This affordance approach in the analysis of design is known as the Function-Behavior-Structure (FBS).

Function of artifacts here is defined as intended behavior. (Wood 2009: 561) In his Computational

Representations of Functions in Engineering Design William H. Wood connects the notions of function

and behavior by defining functions as subsets of behaviors (Wood 2009). Function then is the subset

of behaviors that we design into artifacts. (id.:544) Wood states that this concept is even more

applicable with the modern design of artifacts where more and more use is made of microprocessor

controlled behavior of artifacts. (id.:545)

Affordances can be considered at different levels. This concept has been worked out by Auke Pols in

his chapter on nested affordances. (Pols 2011) These nested affordances can be combined with

different levels of actions in the chain of a functional activity, as mentioned in the previous section.

Joel Norman distinguishes two visual systems that provide an affordance process and relates these

to neurological structures(Norman 2002: 95):

3. The dorsal system deals mainly with the utilization of visual information guiding ones

behaviour in ones environment.

4. The ventral system deals meanly with the utilization of visual information for “knowing” ones

environment.

Activity Theory of Technology (ATT)

The concept of interaction of subject and artifacts also formed the base of the Activity Theory of

Technology (ATT) as a specific approach within the more general paradigm of activity theory.

(Kaptelin and Nardi 2006) This paradigm considers action of actors in the context of culture,

18

motivation, the mediating role of artifacts, social reality and the complexity of real life. Although the

basic concepts of activity theory are similar to phenomenology, it developed, probably independently,

in parallel in the USSR and became known only in the 1980s via Scandinavian researchers. Like the

post-phenomenological view mentioned above, the activity theory also recognizes the mediating role

of artifacts, but this role is not symmetrical as proposed by Verbeek and Latour. One of the

contributions of activity theory of technology is in the attention to the user interface of artifact

design.

Material Engagement Theory (MET)

Much in line with the preceding theories / approaches, based on the concepts of strong interaction

between the mind and the surrounding world, the term Material Engagement is introduced by Colin

Renfrew (2001) and worked out by Lambros Malafouris (2004) in terms of a cognition theory

stressing the point that things shape the mind27. (Renfrew 2009), (Malafouris 2013) .28 With the

archeological profession of Renfrew and Malafouris this Material Engagement Theory concept is

introduced in particular in the archeological analysis of the cognitive development from early

hominines to the current H. sapiens. The authors extend the human interrelation with artifacts of the

concepts above to the domain of evolution.29

Conclusion

Our entanglement with the world and artifacts can be recognized by the way we use mechanical

artifacts, use tools, ride a bicycle, drive a car and so on, while also the use of such things as symbols,

drawings or maps should be included. Today this entanglement also comprises the use of electronic

devices well-demonstrated by the use of blind typing on the keyboard. Another good example of this

entanglement is improvised music, or as a combination of entanglements, a musician composing for

a particular type of instrument directly on paper. The development of this entanglement can be

considered on an individual level by experience and training, but also on a cultural level. For this

later approach the term material culture, as commonly used in archeology, is most applicable.

C2.2.1. Pragmatism and phenomenology with their paradigm of direct, non-propositional, interaction

between subject and artifacts provide a base for the important but more difficult-to-access area of

tacit / partly subconscious entanglement processes in use and design of artifacts in addition to

deliberate activities.

C2.2.2. The concept of affordance is compatible with C.2.2.1.

C2.2.3. The notion of (non-intentional) behavior as attribute to artifacts and material is compatible

with C2.2.1. and C2.2.2.

3.4. Philosophy of mind The interactive paradigm of subjects with their surrounding world is now supported by results from

the neurosciences. The cognitive perspective on interaction with artifacts is based on what is now

27

The title of Malafouris book: How things shape the mind, A theory of Material Engagement. (Malafouris 2013) 28

Malafouris extention to the use of metaphors will be discussed later. 29

As far as I know, the term entangled is not used by these authors, but as I consider it as an appropriate one for this mechanism it will be used further in this context.

19

indicated as embodied cognition. Handling objects integrates the applicable characteristics of the

object in relation to the handling in the neural processes as a kind of inclusion in the body-scheme.

Or, more formally according to the embodiment thesis: many features of cognition are embodied in

that they are deeply dependent upon characteristics of the physical body of an agent, such that the

agent's beyond-the-brain body plays a significant causal role, or a physically constitutive role, in that

agent's cognitive processing. (Wilson and Foglin 2011) The scientific knowledge about this

mechanism substantiates arguments for using the interaction concept in philosophy and supports

the current revival of phenomenology. (Gallagher and Zahavi 2008)

Ahead of current philosophy of mind approaches, Daniel C. Dennett introduced in True Believers a

specific intentional view on artifacts (1971). 30 His intentional stance is based on the observation that

we as humans tend to ascribe intentional attitudes not only to biological subjects but also to

(complex) artifacts. The basic concept seems a valuable observation and is useful for understanding a

basic cognitive mechanism: we intend to interpret artifacts as objects that can behave for a purpose.

He uses the example of the thermostat several times, but adds that assigning behavior is also applied

in daily life with the example of the electrician who explained to him that ‘lightning always wants to

find the best way to ground’. (Dennett 1987:22)

Modeling cognitive aspects of artifacts

The conceptual modeling of Peter Kroes offers a good starting point to integrate a selection of the

above mentioned views.

1. I concur with Kroes’ argument of avoiding the term environment, but I prefer to call it

application context(s), I place it above the artifact’s attributes as the application context is

not directly linked with the artifact itself. On one hand an artifact can be used in different

environments and on the other hand different artifacts can perform the same function in the

same environment. The plural contexts is applicable to cover for the multi-functionality,

concurrent functionality and incidental functionality as mentioned in 3.2.

2. The link between the application context and the artifact can be considered from the two

perspectives of the action and the function as attributes of the artifact, hence the choice in

this paper for functional plan: functional to cover the artifact’s cognitively recognized

attribute and plan (taken from Houkes and Vermaas’ use-plan) to cover the intentional

perspective from the application context with the same considerations as mentioned by

Houkes and Vermaas. This connection also meets Sven Ove Hansson’s comment made on the

introduction of use plan. (Hansson 2006)

3. As the physical structure indeed is a well- defined notion (and will offer a good base for

further analysis in terms of material and form), in a pragmatistic phenomenological approach

it should certainly be considered as a key element.

30

Chapter 2 of his book Intentional stance as a replacement of his article Intentional systems in 1971. (Dennett 1987: 3)

20

4. However, the physical structure does not always give information about how the artifact

performs its function. In Simon’s model this can be expected to be covered by character.31

Based on the various views reviewed in section 2.2. and the use of the term behavior in

section 2.1 the choice for the missing element between the physical structure and the

functional plan is behavior. Knowledge of the behavior is not only important for the user

(think for instance about the limitations in using fire extinguishers) but even more important

for a designer who searches for an improved product. A product with a better performance /

behavior in the same or another application context.

These considerations result in the diagram in figure 3.4 below.

fig. 3.4 Cognitive aspects of artifact

The application context in the diagram can be

understood, depending on the purpose of the analysis,

as the particular context of a user or as the more generic

context of an application area of an artifact to be

considered by the designer.

Behavior should be understood in a plural sense. As

mentioned above: function is the subset of behaviors

that are designed into the artifact. But behavior also

covers not-predefined incidental applications.

Behavior of artifacts and materials does not refer to an ontological concept, but to cognitive

perception that can also be recognized in the linguistic use to characterize materials and artifacts:

Materials: wood keeps warping, a bar bends under lateral pressure, materials shrink or expand

when temperature goes down resp. rises, gasoline explodes in a cylinder of a car by ignition,

electrons accelerate in an electric field, and moving electron rays bend under influence of a

magnetic field, etc.

Artifacts: a thermostat controls temperature, cars behave well or badly at higher speeds,

complex artifacts even take the ”initiative” to breakdown.

Affordance is then to be understood as the knowledge / perception of the functional - or accidental

functional - behavior of an artifact. The functional plan is the plan to apply the behavior. Behavior

covers not only the so-called accidental functions but also the specific aspects of usage in possible

different application environments, such as household versus professional, heavy-duty usage.

31

Kroes does not account for leaving out character, nor does he state whether he includes this in material

structure, but Simon stated that if the artifacts (in his terminology the inner system) is properly designed, its

behavior will be determined in large part by the behavior of the outer environment.(Simon 1996: 11–12)

21

3.5. Conclusions

The main elements of this subsection to be consolidated as applicable input for analysis of actions

with artifacts are:

C3.6.1 A variety of views support the concept of an integrated approach of subjects and their

artifacts. This concept is being confirmed by current neuroscientific results.

C3.6.2 The notion behavior is a common element in the interactive theories of artifacts and is also

used in the descriptions by functionally and intentionally oriented authors.

C3.6.3 The functional and intentional views on artefacts can be considered as: complementary views,

highlighting two ways in which an artefact is related to application. The functional view from the

artefact site, intentional from the application site.

C3.6.4 The diagram of figure 3.4 can be used as a reference model for further analysis.

C3.6.5 Phenomenology, post-phenomenology, the concept of artifact affordance with the function-

behavior-structure and the activity theory of technology all emphasize the interactive aspects of

artifacts. Although some examples can be found of symmetrical interaction between actor and

artifact this cannot be used as a basic concept as proposed by Bruno Latour. Symmetrical interaction

is limited to some guided activities that are designed by humans for a specific purpose. Thus the

symmetrical concept as proposed by Latour and Verbeek seems one step too far.

22

4. Modeling Action with artifacts Based on the analysis and modeling of actions in section 2 and the results of modeling of artifacts

derived from the different views in section 3 in this section the combination of these two are

combined in order to verify the consistency of these two modeling activities. As discussed in section

2 the mainstream action theories are based on agent or state of mind causation, use of an artifact

than starts at the agent site. In relation to artifacts a specific kind of inverse process indicated as

affordance, or ‘what things do’ as mentioned in section 3. This specific process should also covered in

the modeling activities of this section.

As concluded in the section on action theory actions can be understood as multilevel closed loop

systems. The higher levels belong to the domain of practical reasoning the lower levels can be

explained as with control and adaptation closed loop systems.

Practical reasoning

In the process to consider the option for realization of a desire the possibility to apply one or more

artifacts will primarily be based on the functional aspects. But then often also behavioral aspect will

be taken into consideration. I

In the context of this paper esthetical aspects although sometimes even dominant are not included in

this paper because these aspect usually play their role during the acquisition process and not in the

use activity.

Execution

The lower complex of sensory-motor systems develops from childhood by the complex combination

of social learning and own trial and error and error experiences. That is how we learn to master

working artifacts, starting with conscious controlled movement, building up the routine to a

subconscious automatism. By that moment the artifacts behavior is an integrated element in the

neural system. In terms of today’s philosophy of mind: to artifact is embedded.32 As such the artifact

became available as an own capability in the knowledge base of the practical reasoning domain.

fig. 3.5 modeling action with artifact

32

The German resp Dutch expressions: ‘sich etwas aneignen’, ‘zich iets eigen maken’ give a better indication then the English ‘to master’.

23

5. Conclusion

• In relation to artifact use actions are to be

analyzed by

a three main level concept:

o the domain of practical reasoning

o the level of intentional causation

o the sensory-motor body control

level

• ‘Behavior’ is an applicable notion for the

cognitive / action oriented analyses of

materials and artifacts.

• To match multilevel action theory it is

necessary to distinguish between functional

behavior and physical behavior.

• The tripartite view on artifacts and its

relation with application contexts can serve

as reference for further cognitive- and

action-related analyses.

fig.2.1 three level action model

fig. 3.5 modeling action with artifact

24

Appendix 2.1. Explaining versus understanding

The meaning of explanation in several references on action theory varies from a strict scientific

oriented Deductive Nomological (DN) position to the meaning of explanation in daily language that

comes close to understanding. This variation then combines with the meaning of causation, which

ranges from the scientific causation as used in mechanical laws with the example of Hume's billiard

balls to the indication of a cause in daily life.

In today's philosophy of science, as Wesley C. Salmon points out, the Deductive-Nomological model

of Hempel-Oppenheimer is generally considered to be too strict. (Salmon 1972; 20-23, paragraph

problems for the DN pattern of explanation) Scientific understanding is, after all, a complicated affair;

we should not be surprised to learn that it has many different aspects. (Salmon 1972; 20-29)

Salmon refers to three main elements. The first element to be considered, as also identified by Van

Fraassen, is an analysis of the problem under consideration. This should then be followed by the two

other elements, which are the clarification of the underlying mechanisms and the fitting of the

phenomena in a consistent representation of the world. These are the elements also applied in this

thesis.

The following examples indicate some different uses of the terms:

An object in sunlight will warm up and eventually might change color. The object is not acting,

warming up happens to the object. The causality can be understood / explained by basic experience,

more scientifically by macroscopic thermodynamic laws and at low basic physical levels by the energy

of Electromagnetic radiation with absorption at a molecular level of the object.

We see a small tree in partial shade grow in the direction of the sunlight. In daily language we say

the tree is growing to the sunlight because trees need sunlight to grow. We then understand the

reason, and the explanation answers the why question. It does not, however, answers the how

question; when sunlight stimulates growth one would expect the side of the plant with most sunlight

to grow faster with the result that the plant would bend in the direction of the shade, but this is not

what we observe. In this case we explained by reason and we even use law like expressions, but the

why explanation contradicts the how causality.33

Many animals choose their path avoiding exposure in an open area. We explain / understand

this behavior as avoiding the risk of being caught. We understand the reason, although the

behavior of most of the animals will not be based on a rational process.

One might see a person going to a seat in the shade while seats in the sun are also available.

In such a case one can think of different reasons why that choice is made, too warm, allergy

for sunlight, wanting to read a map, expecting a companion who does not like to sit in the

33

The why question has always be understood, It is only since the experiments from 1913 onwards that the causing mechanism could be explained: sunlight generates a plant hormone (auxine) that stretches the cells at the shadow side, the mechanism is called photropism.

25

sun. There is no way of knowing for sure, not even by asking for the reason, for they just

might not be in the mood, but not wanting to say so.

Physicalism and Reductionism

A significant discussion in theory of action is focused on the so-called physicalism issue, and in a

discussion on causality this issue cannot be neglected.

Physicalism claims that the actions of physical objects can be completely explained in terms of

physical causes and laws. The position which assumes that the same mental state is related to the

same physical state is usually called type identity. My position in this paper is that the mental states

such as beliefs and desires are higher level functional identifications of states and processes which in

turn are activations of neurological instances. At a low level certain actions can indeed be explained

by detailed neurological processes (see example in appendix 2). However, many actions of functional

artifact use are far too complex to be completely worked out at a detailed neural level. If this were at

all possible, it would require large computer facilities. But large computer facilities have a similar

problem, as it will not be possible to follow the process at low circuitry, making it impossible to see at

bit level what really happens. Moreover, to perform the same function computers will not act in

exactly the same way at low level. This last remark can easily be demonstrated by a simple example.

This text typed with a particular pc running Microsoft OS would be generated in a different way by

another type pc. This position is usually called token identity. (For some more information and a

survey of references on this issue see e.g. Davis 2010) However, the output string of bytes to the

printer would be identical (type identity), assuming the pc is configured in the right way. It is the so-

called printer driver program that transforms the specific internal codes into the required output.

With human beings we can see this "programming" process taking place. Just after birth infants see

their own arms moving almost at random; gradually they 'program' their neurons to act according to

their will. ("will" is here meant not in an ontological sense, but just as a higher level of action control.)

Another way to address the question of explanation is by referring to the separation in the

vocabulary domains. This is the position also taken by Donald Davidson: the mental is not an

ontological but a conceptual category. (Davidson 1987, 114) Davidson then rejects reductionist

attempts of people like Dretske with an example of black holes. As it is not possible to get

information about the inner structure of black-holes, this lack of information makes is impossible to

complete a theory of the world. 34

34

I think this example is misleading, since the concept of black-holes allows for a complete theory of the observable world only with the exception of the internal structure of the black-hole, because the effect of the black-hole on the observable world is well defined. This is not the way we can isolate functional neural clusters, their effect on the observable world is difficult to define.

26

Appendix 2.2. Feedback loops in action theory

Concept and mathematical basis

The basic concept of a feedback loop control is the structure by which the actual value of a process

indicator (R), for instance a temperature, is compared with a desired value (D). The difference

between these two values is used to control the process.

Comparison of Dretske's diagram with usual feedback loop:

The two figures below show the difference between the usual way to explain the concepts and

Dretske's diagram. (Dretske 1988: 88)

Fig A-2.2.1. comparison Dretske's model with usual feedback schematics

In the feedback, the measured value of the process characteristic is subtracted from the desired

value and the difference is used to drive the adjustment of the process as indicated in figure A2.2

below.

Both Davidson and Dretske refer to the example of a thermostat, but they do not refer to the basic

concept of feedback loop systems. Dretske mentions the bi-metal strip that forms the heart of old

thermostats. The disadvantage of this example is that it is an ingenious combination of a measuring,

reference and switch function which makes it not ideal to understand the basic concept.

Servo-concept

A more specific concept structure is the so-called servo loop. These processes can also be of a

mechanical nature as used in factory mechanization and robotics. The basic structure is the same,

but the difference is that the desired value is the variable. By changing the variable, the system

realizes the desired value, also when environmental conditions (within some limitations) change. This

structure is shown in figure A2.2.

27

Fig. A-2.2.2 Mathematical basis of feedback loop concept

The key characteristic again is that the difference between the desired and the actual value is used to

realize a result. The control algorithm supplies the reference for the force that realizes the required

result.

Feed Back Loops in neural systems

Already in the 50s of the last century McGulloch and Pitt showed that there exists feedback-loop

configuration of neurons in the brain. (Pask 1961) In practice, in this kind of neuron structure several

feedback loop configurations work together so that one configuration can adapt the performance of

another loop, which can in turn correct another one.

The Vestibulo Oculair Reflex is a prototype

example. The eye muscles make corrections for

the movement of the head in order to stabilize the

projection on the retina. This control mechanism is

handled by a group of neuron cells in the

cerebellum. These cells combine the inputs from

the organ of balance, the eye muscles and the

retina and transmit these to a special group of

cells. These cells control the action. When the

feedback is insufficient or too strong, the shift on

the retina is sent to a third group of cells (the

lower olivecernels), and these send a correction

signal to the Purkinjecells which adjust the

feedback mechanism.

fig. A-2.2..3 multilevel nested feedback loops

When we extrapolate these structures to the processes in the brain in general it becomes clear that

one cannot speak of states in the brain as entities, but that we should consider these as complex

dynamic systems that can also be understood and explained by a final cause explanation.

It should also be clear from the above explanations that the specific property of the feedback-loop

configuration cannot be reduced to a lower physical level. This observation is compatible with the

more generic metaphysics of Lynne Rudder Baker. (Baker 2007: Chap. Commonsense causation)

28

With the knowledge of feedback processes, we can pick up Davidson's example to explain how a

mental desire can result in a bodily action. We start at the stage where the practical reasoning as

mentioned above makes someone decided to call for a taxi. The person in question with a paralyzed

arm and carrying something with his other arm, is equipped with a robotic support-arm for the

paralyzed arm which has a voice recognition device. This starts the servo system of the robotic arm

by the command to raise the arm when the vocal `taxi' is detected. As Davidson states, I take it as

obvious that linguistic behavior is intentional, we can explain that the intention caused the action.

Assuming the similarity of the earlier described neural systems is accepted, this would also apply to

non-paralyzed arms without robotic support.

More complex systems

The above-mentioned example of raising the arm covers only a well-defined, relatively low level

action process, starting with an already available mental desire to raise an arm since that is typical

for calling a taxi. The desire has been made well-defined as an intention by the linguistic interface to

avoid subconscious processes that might have played a role in the decision process to call for a taxi

rather than to walk or use public transport. That kind of decision process requires another type of

analysis.

This not fully reductionist approach is also advocated by Alicia Juarrero. In her Dynamics in action,

Intentional Behavior as a Complex System she analyses what the Dynamic Systems Theory implies for

the philosophy of mind and the philosophy of action. (Juarrero 1999) The General Dynamic Systems

Theory, first articulated in 1981 by the biologist Ludwich von Bertanlanffy, is applicable to systems

where the parts and the whole interact. This means subsystems contribute to the system as a whole

and receive information from other subsystems and / or the higher levels of the system. (id.: chapter

7) Juarrero concludes that once a semantic state space self-organizes, the trajectories within that

higher dimensional, neurological space embody new rules: the constraints of logic, meaning, value

and the like. (id.:248)

It is interesting to notice that Salmon, in the quote mentioned above, moved from a justification of

explanation to understanding. At the level of basic mechanisms final cause can be used as a

justifiable explanation. In case of more complex systems criteria for the label justified explanation

should be developed. Gradually one will move into understanding.

29

References

References for section 2

Allen, Colin. 2003. “Teleological Notions in Biology.” Stanford Encyclopedia of Philosophy. Stanford

CA: Metaphysics Research Lab, CSLI, Stanford University. Fil/cogn fil/ causes. y.

Audi, Robert. 1997. “Acting for Reasons in \cite Mele 1997.” In The Philosophy of Action. Oxford

University Press (1997/2003),.

Baker, Lynne Rudder. 2007. The Metaphysics of Everyday Life: An Essay in Practical Realism.

Cambridge: Cambridge University Press.

Davidson, Donald. 1987. “Problems in the Explanation of Action in Problems of Rationality.” In

Problems of Rationality, edited by Donald Davidson, 101–16. Clarendon Press Oxford (2004).

Davis, Wayne A. 2010. “The Causal Theory of Action.” In A Companion to the Philosophy of Action.

Wiley- Blackwell.

Dennett, Daniel C. 1987. The Intentional Stance. MIT press, Cambridge MA.

Dretske, Fred. 1988. Explaining Behavior. MIT press (1988).

Futó, Judit, Erno Téglás, Gergely Csibra, and György Gergely. 2010. “Communicative Function

Demonstration Induced Kind-Based Artifact Representation in Preverbal Infants.” Cognition 117: 1–8.

Gallagher, Shaun, and Dan Zahavi. 2008. The Phenomenological Mind, an Introduction to the

Philosophy of Mind and Cognitive Science. Routledge, London.

Hansson, Sven Ove. 2006. “Defining Technical Function.” Studies in History and Philosophy of Science

Part A 37 (1): 19–22. doi:10.1016/j.shpsa.2005.12.003.

Harman, Gilbert. 1997. “Practical Reasoning.” In The Philosophy of Action, Ed Alfred R Mele, 149–77.

Oxford Readings in Philosophy. Oxford University Press (1997/2003),.

Hornsby, Jennifer. 2011. “Actions in Their Circumstances.” In Essays on Anscombe’s Intention.

Harvard University Press, Cambridge Massachusetts.

Houkes, Wybo, and Pieter E. Vermaas. 2010. Technical Functions, on the Use and Design of Artifacts.

Springer, Dordrecht.

Hughes, Jesse. 2009. “Practical Reasoning and Engineering.” In The Handbook of The Philosophy of

Technology and Engineering Sciences.

Ihde, Don. 1983. Existential Technics. State of New York University Press, Albany, NY.

30

———. 1991. Instrumental Realism: The Interface between Philosophy of Science and Philosophy of

Technology. The Indiana Series in the Philosophy of Technology. Bloomington: Indiana University

Press.

Juarrero, Alicia. 1999. Dynamics in Action, Intentional Behavior as a Complex System. MIT press

(edition 2002).

Kaptelinin, Victor, and Bonnie A. Nardi. 2006. Acting with Technology. MIT press, Cambridge MA.

Kroes, Peter. 2006. “Coherence of Structural and Functional Description of Technical Artefacts.”

Studies in History and Philosophy of Science 37: 137–51.

———. 2012. Technical Artefacts: Creations of Mind and Matter. New York: Springer.

Kroes, Peter, and Antonie Meijers. 2006. “The Dual Nature of Technical Artefacts.” Studies in History

and Philosophy of Science 37: 1–4.

Malafouris, Lambros. 2013. How Things Shape the Mind: A Theory of Material Engagement.

Mele, Alfred R. 1997. “Introduction.” In The Philosophy of Action. Oxford University Press,Oxford.

Millgram, Elijah. 2012. “Practical Reason and the Structure of Actions.” In The Stanford Encyclopedia

of Philosophy. Vol. 2012. url{http://plato.stanford.edu/archives/sum2012/entries/practical-reason-

action/}},.

Mitcham, Carl. 2006. “From Phenomenology to Pragmatism: Using Technology as an Instrument.” In

Postphenomenology, A Critical Companion to Ihde. State University of New York Press.

Neander, Karen. 1993. “Misrepresentation and Malfunctioning.” Philosophical Studies, no. 79-2: 109–

41. Filosofie van de techniek.

Norman, Joel. 2002. “Two Visual Systems and Two Theories of Perception: An Attempt to Reconcile

the Constructivist and Ecological Approaches.” Behavioral and Brain Sciences 25: 73–96.

Pacherie, Elisabeth. 2008. “The Phenomenology of Action: A Conceptual Framework.” Cognition 107

(1): 179–217. doi:10.1016/j.cognition.2007.09.003.

Pask, Gordon. 1961. An Approach to Cybernetics, Vert. Inleiding Tot de Cybernetica. Het Spectrum

1965.

Pols, Auke J. K. 2011. “Acting with Artefacts”. Technical University Eindhoven.

Preston, Beth. 2009. “Philosophical Theories of Artifact Function.” In The Handbook of The Philosophy

of Technology and Engineering Sciences. Elsevier, Amsterdam.

Renfrew, Colin. 2009. Prehistory: The Making of the Human Mind. Modern library paperback ed. A

Modern Library Chronicles Book v. 30. New York: Modern Library.

Salmon, Wesley. 1972. “Philosophy of Science.” In , edited by Wesley Salmon eo, 7–41. Hackett

Publishing Comp, Indianapolis/Cambridge.

31

Simon, Herbert A. 1996. The Sciences of the Artificial. Cambridge, Mass.: MIT Press.

Stoutland, Frederick. 2011. “Anscombe’s Intention in Context.” In Essays on Anscombe’s Intention.

Havard UP.

Tromp, Hans H. M. 2013. “Design View on Functional Artifacts, in Search of the Input for a Cognitive

Analysis of Artifact Design.” In , SPT2013 conferenceTrack 11. Lisbon.

Verbeek, Peter-Paul. 2000. What Things Do. Pennsylvania State Univerity Press, (2005).

Wilson, Robert A., and Lucia Foglin. 2011. “Embodied Cognition.” Standford Encyclopedia of

Philosophy. http://plato.stanford.edu/entries/embodied-cognition/#Phe.

Wittgenstein, Ludwig. 1945. Philosophical Investigations I. Edited by G. E. M. Anscome eo. Blackwell,

Oxford 1958.

Wood, William H. 2009. “Computational Representations of Function in Engineering Design.” In

Philosophy of Technology and Engineering Sciences. Elsevier, Amsterdam.

References for section 3

Allen, Colin. 2003. “Teleological Notions in Biology.” Stanford Encyclopedia of Philosophy. Stanford

CA: Metaphysics Research Lab, CSLI, Stanford University. Fil/cogn fil/ causes. y.

Audi, Robert. 1997. “Acting for Reasons in \cite Mele 1997.” In The Philosophy of Action. Oxford

University Press (1997/2003),.

Baker, Lynne Rudder. 2007. The Metaphysics of Everyday Life: An Essay in Practical Realism.

Cambridge: Cambridge University Press.

Davidson, Donald. 1987. “Problems in the Explanation of Action in Problems of Rationality.” In

Problems of Rationality, edited by Donald Davidson, 101–16. Clarendon Press Oxford (2004).

Davis, Wayne A. 2010. “The Causal Theory of Action.” In A Companion to the Philosophy of Action.

Wiley- Blackwell.

Dennett, Daniel C. 1987. The Intentional Stance. MIT press, Cambridge MA.

Dretske, Fred. 1988. Explaining Behavior. MIT press (1988).

Futó, Judit, Erno Téglás, Gergely Csibra, and György Gergely. 2010. “Communicative Function

Demonstration Induced Kind-Based Artifact Representation in Preverbal Infants.” Cognition 117: 1–8.

Gallagher, Shaun, and Dan Zahavi. 2008. The Phenomenological Mind, an Introduction to the

Philosophy of Mind and Cognitive Science. Routledge, London.

32

Hansson, Sven Ove. 2006. “Defining Technical Function.” Studies in History and Philosophy of Science

Part A 37 (1): 19–22. doi:10.1016/j.shpsa.2005.12.003.

Harman, Gilbert. 1997. “Practical Reasoning.” In The Philosophy of Action, Ed Alfred R Mele, 149–77.

Oxford Readings in Philosophy. Oxford University Press (1997/2003),.

Hornsby, Jennifer. 2011. “Actions in Their Circumstances.” In Essays on Anscombe’s Intention.

Harvard University Press, Cambridge Massachusetts.

Houkes, Wybo, and Pieter E. Vermaas. 2010. Technical Functions, on the Use and Design of Artifacts.

Springer, Dordrecht.

Hughes, Jesse. 2009. “Practical Reasoning and Engineering.” In The Handbook of The Philosophy of

Technology and Engineering Sciences.

Ihde, Don. 1983. Existential Technics. State of New York University Press, Albany, NY.

———. 1991. Instrumental Realism: The Interface between Philosophy of Science and Philosophy of

Technology. The Indiana Series in the Philosophy of Technology. Bloomington: Indiana University

Press.

Juarrero, Alicia. 1999. Dynamics in Action, Intentional Behavior as a Complex System. MIT press

(edition 2002).

Kaptelinin, Victor, and Bonnie A. Nardi. 2006. Acting with Technology. MIT press, Cambridge MA.

Kroes, Peter. 2006. “Coherence of Structural and Functional Description of Technical Artefacts.”

Studies in History and Philosophy of Science 37: 137–51.

———. 2012. Technical Artefacts: Creations of Mind and Matter. New York: Springer.

Kroes, Peter, and Antonie Meijers. 2006. “The Dual Nature of Technical Artefacts.” Studies in History

and Philosophy of Science 37: 1–4.

Malafouris, Lambros. 2013. How Things Shape the Mind: A Theory of Material Engagement.

Mele, Alfred R. 1997. “Introduction.” In The Philosophy of Action. Oxford University Press,Oxford.

Millgram, Elijah. 2012. “Practical Reason and the Structure of Actions.” In The Stanford Encyclopedia

of Philosophy. Vol. 2012. url{http://plato.stanford.edu/archives/sum2012/entries/practical-reason-

action/}},.

Mitcham, Carl. 2006. “From Phenomenology to Pragmatism: Using Technology as an Instrument.” In

Postphenomenology, A Critical Companion to Ihde. State University of New York Press.

Neander, Karen. 1993. “Misrepresentation and Malfunctioning.” Philosophical Studies, no. 79-2: 109–

41. Filosofie van de techniek.

Norman, Joel. 2002. “Two Visual Systems and Two Theories of Perception: An Attempt to Reconcile

the Constructivist and Ecological Approaches.” Behavioral and Brain Sciences 25: 73–96.

33

Pacherie, Elisabeth. 2008. “The Phenomenology of Action: A Conceptual Framework.” Cognition 107

(1): 179–217. doi:10.1016/j.cognition.2007.09.003.

Pask, Gordon. 1961. An Approach to Cybernetics, Vert. Inleiding Tot de Cybernetica. Het Spectrum

1965.

Pols, Auke J. K. 2011. “Acting with Artefacts”. Technical University Eindhoven.

Preston, Beth. 2009. “Philosophical Theories of Artifact Function.” In The Handbook of The Philosophy

of Technology and Engineering Sciences. Elsevier, Amsterdam.

Renfrew, Colin. 2009. Prehistory: The Making of the Human Mind. Modern library paperback ed. A

Modern Library Chronicles Book v. 30. New York: Modern Library.

Salmon, Wesley. 1972. “Philosophy of Science.” In , edited by Wesley Salmon eo, 7–41. Hackett

Publishing Comp, Indianapolis/Cambridge.

Simon, Herbert A. 1996. The Sciences of the Artificial. Cambridge, Mass.: MIT Press.

Stoutland, Frederick. 2011. “Anscombe’s Intention in Context.” In Essays on Anscombe’s Intention.

Havard UP.

Tromp, Hans H. M. 2013. “Design View on Functional Artifacts, in Search of the Input for a Cognitive

Analysis of Artifact Design.” In , SPT2013 conferenceTrack 11. Lisbon.

Verbeek, Peter-Paul. 2000. What Things Do. Pennsylvania State Univerity Press, (2005).

Wilson, Robert A., and Lucia Foglin. 2011. “Embodied Cognition.” Standford Encyclopedia of

Philosophy. http://plato.stanford.edu/entries/embodied-cognition/#Phe.

Wittgenstein, Ludwig. 1945. Philosophical Investigations I. Edited by G. E. M. Anscome eo. Blackwell,

Oxford 1958.

Wood, William H. 2009. “Computational Representations of Function in Engineering Design.” In

Philosophy of Technology and Engineering Sciences. Elsevier, Amsterdam.