Lecture 6

Question 1

Explaining individual acts

Instrumental rationality:

Answer 1

= to provide reasons for action
Three important issues:
1. How we explain behaviour
2. The role of instrumentalism in social science
3. Differentiating agents from actors
➔ Instrumental rationality focusses on agents as an abstract concept
2. Explaining individual acts
E.g., the happiness of young parents (de Bruin)
Q: How to explain individual choices with regard to having a child?
- Through instrumental rationality (“explanation”)?
- Then you expect parents to make lists of pros and cons: what makes you happy and
what hinders you?
o Individuals are then seen as agents who make assessments that can be
expressed in “laws”.
o Recall logical positivism / Popper hypothetical-deductive method
But is this the whole story?
- Or is there some sort of “basic feeling” involved, connected with values and norms
that are embedded in the communities where young parents belong to?
- It’s not just a matter of rational consideration; it is a meaningful act of actors that
takes place in a certain context formed by rules.
o Recall Schutz, Taylor, Geertz:
o Common-sense thinking (of the subjects at group level)
o Social scientific models of motivations, feelings, meanings

Question 2

Hempel’s Covering Law Model (CLM)

Answer 2

• There is a phenomenon A that must be explained:
  o The explanandum (“what has to be explained”) aka
  the effect.
  o = what is to be explained (effect)
• How? Explain A by specifying (i) a general law and (ii) an
  initial condition:
  o The explanans (“that which explains”)
  o = what does the explaining (law = cause)
• The explanans in the CLM need to have a law: otherwise it
  isn’t a scientific explanation and it wouldn’t fit in to the
  positivist worldview.
  According to Hempel: “(i) a law and (ii) the initial conditions logically
  entail the event A to be explained.”
  CLM is another name for the deductive-nomological method
• (the hypothetical-deductive method is one variation of that)
  E.g., of CLM: ‘broken bottles
• ’On Monday morning we find bottles left outside are broken (A).
  o (A) must be explained and is therefore the explanandum.
  General law: if the temperature is below a certain value, glass bottles will break.
  Initial condition: it has frozen hard enough on Sunday night to shatter the bottles.
• Phenomenon (A) can be explained on the basis of (i) general law and (ii) this
  specific initial condition

Question 3

.

Answer 3

.

Question 4

Hempel

Answer 4

In this model, the outbreak of WWI is explained in the same way as the shattering of the
bottles.
- In both cases, the method to explain by laws is limited to general aspects of the
events to be explained.
- In other words, intentional actions also belong to the phenomena explained by the
CLM.
- According to Hempel, a reason is a special kind of cause.
For the explanans to successfully explain the explanandum several conditions must be met
Deductive conditions:
- The explanandum must be a logical consequence of the explanans and, the sentences
constituting the explanans must be true
- Nomological (lawful) condition
- The explanans must contain at least one “law of nature” (more info on the next slide)
- This must be an essential premise in the derivation in a sense that: the derivation of
the explanandum would not be valid if this premise were removed

Question 5

Causes and reasons

The ‘I’ –two perspectives: a paradox?

Answer 5

Recall discussion of Descartes and the subject-object distinction:
➢ The ‘I’ as embodied subject: an observable object of knowledge, a material being.
➢ The ‘I’ as pure reason: the perspective under which I experience myself, the elusive
center of my self-awareness.
Descartes could not reconcile this dualism
Positivists tried to reduce/explain away the mind/subject (there is just this physical world)

Question 6

How can a person act freely in a world where everything is subjected to the law of cause
and effect?
According to Kant, this is not a contradiction but a paradox, because people appear under
two different perspectives:

Answer 6

• The material body is subject to the laws of nature in which causal relationships take
  central stage.
• Our mentality -apart from the laws of nature -is a free being, where causes give way
  to reasons.
  Causes and reasons are not the same; they refer to different aspects of reality!
  Kant: don’t substitute causes for reasons, and vice versa!
  • Causes -> empirical regularities (about, for example, exploding bottles at a certain
  temperature) and predictions. No reference is required to motives and meanings.
  • Reasons -> principles of action (which always refer to intentions), psychological and
  social regularities (exceptions possible). Reasons are subjective.
  Reasons are related to rules of conduct that can be violated, while causal laws always apply
  (Winch/Risjord).
  2 level of reasons:
  Hempel: CLM considers ‘intentions’ and ‘reasons’ as causes that refer to general laws.
• Winch: reasons are not causes; reasons refer to socially determined rules.
  Davidson: what is the primary reason for an action?
• The primary reason for an action is always described in causal terms!
• Therefore, “the primary reason for an action is its cause.”
• But such causes do NOT refer to laws!

Question 7

Hempel, Winch and Davidson on reasons

Answer 7

Hempel: CLM considers ‘intentions’ and ‘reasons’ as causes that refer to general laws.
Winch: reasons are not causes; reasons refer to socially determined rules.
Davidson: what is the primary reason for an action?
- The primary reason for an action is always described in causal terms!
- Therefore, “the primary reason for an action is its cause.”
- But such causes do NOT refer to laws!
- Synthesizes aspects of Hempel / Winch
Davidson’s theory of action is closely tied to rational choice theory
- Uses tools of folk psychology (belief and desires) to understand reasons
- But Davidson is not committed to instrumental rationality

Question 8

Rational Choice Theory (RCT)

Answer 8

Goals and beliefs are translated into preferences.
- Preferences are translated into utility functions that organize preferences from high
to low.
- Preferences are expressed in values (ordinal numbers) and the optimum choice is
calculated.
RCT main assumptions
Humans are perfect rational beings
- Their preferences are completely ordered by priority.
- They have complete information about choices and outcomes.
- They can perfectly determine the optimum choice.
Connections with the notion of instrumental rationality
- RCT is about instrumental rationality
o Preferences are the translation of supposed motives and reasons.
o No empirical research is conducted into the actual reasons that people have
for their behaviour.
RCT: decision theory, game theory, social choice theory, and other micro-economic
applications

Question 9

Dominant strategy

Answer 9

- An optimal strategy for a player, 
without being dependent on the 
strategy of other players
Nash equilibrium
- A situation nobody can do better by changing his choice

Question 10

Dominant stratedy and nash equilibrium

Answer 10

Kenney has no dominant strategy:
- There is no optimal strategy that is not dependent on the strategy of Imamura.
Imamura has a dominant strategy:
- Always play ‘north’
There is one “Nash equilibrium”:
- Imamura and Kennedy play North:
- Nobody can do better by changing their choice unilaterally.

Question 11

Challenges for rational choice theory:

Answer 11

Are human beings rational? Is the underlying view of man plausible?
➢ Balancing preferences presupposes the ideal of (psychological) autonomy. Is this
tenable?
➢ Too much focus on an “I-centered conception of agency”?
➢ Sometimes the rational act is not the right act

Question 12

Preferences, autonomy, and rationality
Q: Do we have all the needed information to make a choice?
Example: selecting a healthcare plan

Answer 12

This assumes we (as rational agents) can:
➢ Acquire information
➢ Understand information
➢ Know how to value information
➢ Make a choice based on information
➢ Act in accordance with preferences formed..

Question 13

Sometimes the rational act is not the right act

Answer 13

RCT: It is always avbout my preferences. My preferences do not necessarily have to be
selfish!
Sen’s Distinction between:
- Sympathy: one’s interests are acknowledged, and these interests correspond to your
own interests.

• Commitment: you confirm one’s choice, even if this goes against your own interest.
  Against RCT: we need to quantify the utilities in a better way!
  But behavioral experiments (e.g., Guala) show that motivation is complicated…

Question 14

Chapter 5 book – Action and Agency
Explaining Action
Admiral Tryon and Instrumental rationality

Answer 14

Understanding the motivations of historical figures is often an important aspect of
understanding the events of which they were part
E.g., → Vice-Admiral George Tryon, commander of the British Mediterranean fleet, ordered
two iron-clad battleships equipped with rams to turn toward each other. They collided,
sinking the Victoria. What was he thinking?
➢ Many hypotheses why he did what he did
No one will ever know for sure what Tryon’s intentions were when he gave the order which
resulted in the sinking of the Victoria. Individual theories can start and end with the possible
interpretations of those five words he spoke on the chart house: “It was all my fault”
➢ As a first pass at the idea of an action explanation we might say that the explanation
of an intentional action presents the agent’s reasons, and these reasons make the
action instrumentally rational. Intentional action explanations are not causal because
a causal explanation does not assume that its subject is instrumentally rational.
➢ With respect to naturalism, the concern with agents and actions means that the
social sciences require a different form of explanation than the natural sciences. With
respect to reductionism, the concern with intentional action means that there is
something special about the level of agency.

Question 15

The Function of General Laws in History

Answer 15

Any attempt to contrast the forms of explanation in the natural sciences with the social
sciences requires some conception of “explanation.”
→ The Victoria sank because water entered the ship, which made it at some point
impossible to stay floating = General law (fits into lawful pattern or regularity)

Question 16

Carl Hempel: a scientific explanation must have three features

Answer 16

Carl Hempel: a scientific explanation must have three features
Scientific laws express generalizations by relating properties or types of events.
➢ Hempel drew the conclusion that an explanation never explains the entirety of an
individual event. It explains only the particular properties of the event described by
the law.
Gedownload door: devijabiere | devijaaa.b@gmail.com
Dit document is auteursrechtelijk beschermd, het verspreiden van dit document is strafbaar.
Stuvia - Koop en Verkoop de Beste Samenvattingen
E.g. Hempel:
1. If a person, S, wants to achieve goal G, and believes that doing action A is the best
way to achieve G, then S will do A.
2. S wants to achieve G.
Therefore, S does A.
1. = nothing more than a well-confirmed, empirical law of human behavior
2. = an initial condition
→ Together they entail the description of the event to be explained, intentional action
explanations fit the Deductive-Nomological pattern of explanation

Question 17

Reasons and Causes
Two differences between reasons and causes that suggest it’s mistaken to treat the principle
of instrumental rationality as a causal law

Answer 17

1. It invokes rationality, the principle of instrumental rationality is normative. It says
   what an agent ought to do, not what he or she will do
2. In the social sciences, the subjects of our inquiry have their own language, and they
   have ways of thinking and talking about their actions. Causal laws need make no
   reference to meaning or motive.

→ Peter Winch used these two differences to argue that human action cannot be explained
by laws
➢ Treating the principle of instrumental rationality as if it were a causal law is to
misunderstand the very idea of a reason for action. The principle of instrumental
rationality is normative.

Question 18

Response to Winch → reasons could not be causes: Donald Davidson

Answer 18

Donald Davidson pointed out that while
there may be many reasons that would justify an agent’s action, only one of these will be the
reason why he or she acted. → primary reason
➢ Without some kind of causal relationship between belief, goal, and action, then there
is nothing to make the primary reason primary. → primary reason = cause
➢ He holds therefore that an explanation needs to be causal
➢ It is entirely possible for S to want G, believe that A is the best way to achieve A, and
yet fail to do A.
o While primary reasons are causes, action explanation does not follow the
Deductive-Nomological pattern of explanation.
→ Davidson disagrees with both Hempel and Winch

Question 19

Re-enactment: Verstehen Revisited

Answer 19

Verstehen: that social scientific understanding must capture the meaning the events have
for the subjects.
To properly understand an agent’s reasons in their capacity to move the agent to action, the
action needs to be understood subjectively, from the inside.

Question 20

Collingwood’s re-enactment

Answer 20

➢ There is an ineliminable use of the historian’s own ability to act for a reason.
➢ The historian must imagine the situation facing an agent, including the environment,
the costs and benefits of different courses of action, the reactions of others, and so
on.
➢ Then the historian thinks the problem through for him or herself.
Recent revival in the light of new psychological and philosophical theories
➢ The “theory-theory” holds that children learn about mental states in the same way as
they learn about any other part of the world.
➢ My ability to reason about what to do in the light of my own beliefs and attitudes
produces conclusions about what I would do in that situation. I thereby understand
why you acted as you did.
One challenge for a re-enactment view of social scientific understanding is that it needs to
find an appropriate role for evidence, historical or other- wise.
➢ One might challenge the re-enactment view, by arguing that re-enactment is not
sufficient for historical (and other social scientific) understanding.
➢ In response, most proponents of simulation or re-enactment agree that theorizing of
the ordinary sort is also necessary for social scientific under- standing.

Question 21

The Games People Play

Answer 21

The principle of instrumental rationality is at the center of one of the most powerful
explanatory paradigms in the social sciences. “Rational choice theory” or “RCT”
➢ It uses instrumental rationality to explain recurring patterns of interaction.
➢ It is a body of definitions and constructs that are useful for modeling a variety of
social phenomena.

Question 22

Rationality and Utility

Answer 22

“If a person has goal G, and believes that doing A is the best means to achieve G, then the
person will do A.”
➢ Ignores that people have many goals that are sometimes inconsistent
➢ Different means to a particular goal may be more or less likely to succeed, “best
means” raises a problem of ranking
The problem of ranking goals and means is made more difficult by the fact that they interact.
➢ The strength of different desires interacts with the probability of success. One might
reasonably prefer a sure path to a less desirable goal, to a risky path to a more
valuable one

Question 23

Utility functions:

Answer 23

a rational agent doesn’t just have a “goal G”, he or she has preferences
among all the available functions
➢ Ordinal utility functions: ranks a set of possible objects highest to low
o His preferences must be transitive: If he prefers x to y, and y to z, then he
must prefer x to z

Question 24

Games and Strategies

Answer 24

Putting the game into this matrix form helps us think
through the strate- gies from each player’s point of view.
➢ Where the player has a strategy that brings him
higher utility than any other strategy, no matter
how the other agents act, the player is said to have
a dominant strategy
Equilibrium: set of strategies composed from the best strategy for each player

→ if all strategies in the set are dominant strategies = dominance equilibrium
→ Well known game with unique dominance equilibrium = prisoners dilemma
→ Dominance equilibrium in figure down below = {confess, confess}

→ if all strategies in the set are dominant strategies = dominance equilibrium
→ Well known game with unique dominance equilibrium = prisoners dilemma
→ Dominance equilibrium in figure down below = {confess, confess}

Question 25

→ if all strategies in the set are dominant strategies = dominance equilibrium
→ Well known game with unique dominance equilibrium = prisoners dilemma
→ Dominance equilibrium in figure down below = {confess, confess}

Answer 25

➢ When the moves are simultaneous, there is no rational way to achieve coordination.
➢ Choosing sequentially resolves the dilemma

Question 26

Agency

The Psychological Plausibility of Rational Choice Theory

Answer 26

Some philosophers and social scientists have argued that rational choice theory is
psychologically implausible and concluded that it is of little use in social scientific
explanation.
Rational choice theorists recognize that decision theory and game theory do not do justice
to the richness of human motivations
Two complains about psychological implausibility
➢ Defenders will argue that decision theory and game theory are not meant to be
complete theories of human behavior.
➢ RCT explains typical human behavior, not the actual behavior of a specific individual.

Question 27

Amartya Sen argued

Answer 27

Amartya Sen argued that conceptualizing motivation in terms of utility maximization
conflates two different kinds of reason for action. Distinguished “sympathy”-“commitment.”
E.g.; two boys who find two apples, one large, one small. Boy A tells boy B, “You choose.” B
immediately picks the larger apple. A is upset and permits himself the remark that this was
grossly unfair. “Why?” asks B. “Which one would you have chosen, if you were to choose
rather than me?” “The smaller one, of course,” A replies. B is now triumphant: “Then what
are you complaining about? That’s the one you’ve got!”
➢ In this story, B understands all motivations in terms of “sympathy.” Each individual
has particular desires, and acts to obtain those desires. As a result, B interprets A as
desiring the smaller apple.
➢ Choices based on commitment, according to Sen, may run contrary to self-interest.
Game Theory in the Laboratory (Guala’s experiment)

The fact that some subjects do not choose the “defect” strategy shows that their utilities are
not limited to the monetary payoffs offered by the experimenter. And when the payoffs
change, the game changes.
➢ The possibility of refining utility functions raises a question about the empirical
testability of rational choice theory.
➢ Whatever an agent does must therefore have had the highest utility at the time.
Guala’s argument supports Sen’s view that rational choice theory inappropriately treats
agents as egoistic. Not all motives can be understood simply as a stronger preference for
one outcome over another.
Conclusion: that the limitations of the theory prevent us from treating it as a fully general
theory of action

Question 28

Instrumentalism and Structuralism

Answer 28

Rational choice theory is important precisely because it is able to travel among different
fields and to link them together. Such cross-fertilization is difficult to understand on an
instrumentalist construal of science.
The way that rational choice theory is used in practice has led some to interpret it as a
“structural” theory. The model postulates idealized agents who are responding to specific
costs and benefits.
Treating rational choice theory as a structural theory distances rational choice explanations
from the explanation of (individual) intentional action.