SARAH PAUL SEPT 28, 2023

3. ACTION
EXPLANATION
 2

PREFACE

▸ 1. (例 ) Imagine that you walk into a room and see your friend
Mariko peel a banana, put the peel into a blender, and throw the
banana in the garbage. You ask her “Why are you doing that?”

(回答 ) One possibility is that she might look in the blender and
exclaim “Oops, that’s not what I meant to do – I wasn’t paying
attention!”
This answer seems to indicate that the action of blending the peel
and discarding the banana was not intentional in the sense that
we philosophers aim to understand.
一
子
 3

▸ (回答 ) A second possibility is that she replies “I’m making a low-

calorie, high- ber smoothie! I want to get more ber in my diet,
and I read that all the calories are in the banana and all the ber is
in the peel.”
This explanation implies that Mariko is blending the peel
intentionally.

Whereas the rst explanation merely identi es the cause of her

happening to throw away a good banana – a lack of attention – the
second one identi es a reason for throwing it away. Call this
second kind of reply to the “Why?” question a “rationalizing”
explanation (合理化解釋，理由解釋).
二
fi
fi
fi
fi
fi
fi
 4

▸ 2. Rationalizing explanations of action reveal the light in which the

agent viewed the action as worth doing, or something that it
would make sense to do. ( 動的合理化解釋揭 了某個 ( 動主體
的) 看法： 動主體視該 動為值得採取、或是某件合理可 的事。)
In Donald Davidson’s words, the explanation “... leads us to see
something the agent saw, or thought he saw, in his action – some
feature, consequence, or aspect of the action the agent wanted,
desired, prized, held dear, thought dutiful, bene cial, obligatory,
or agreeable” (1963, 685).
行
行
行
示
fi
行
行
 5

▸ (回答 提供了 個合理化解釋) In this case, Mariko was making a

banana-peel smoothie because she thought it would be healthy
and slimming and therefore good, or at least intelligible. Of
course, she may be mistaken about this; what matters for
explaining her action is that she thinks there is a good reason for
doing it.
(回答 是 個因果解釋，但它不是 個理由解釋) In contrast,
although a lack of attention explains why she threw the banana
away in the rst case, it is not the kind of thing that could justify
her doing so. Though we might say that inattentiveness is the
“reason why” she threw it away, this sense of “reason” refers to a
cause rather than a consideration that shows what happened to
be good or rationally intelligible.
二
一
一
fi
一
一
 6

▸ 3. 為什麼本書作者先討論 動解釋 (action explanation) 不是先問

動是什麼？

… many philosophers follow G.E. M. Anscombe in thinking that

rationalizing explanations are central to understanding the kind
of agency we should be interested in: actions that are done for
reasons, and therefore subject to the sense of the question
“Why?” that aims to elicit those reasons.
行
行
而
 7

▸ 4. So: how exactly do the reasons cited in answer to the question

“Why?” explain what an agent is doing or has done?
 8

3.1 GUISES OF RATIONALIZING EXPLANATION

▸ 動解釋有幾種形式：

▸ 1. 引述理由解釋 (引述 動主體為什麼採取該 動的理由)：

First, some rationalizing explanations explicitly cite a
consideration that was the agent’s reason for acting:

Reasons: “Mariko’s reason for eating a banana peel is/was that

eating brous things is good for the digestion.”
行
fi
行
行
 9

▸ Reason-citing explanations pick out some feature of the action

that justi es it or otherwise renders it intelligible, at least from the
agent’s perspective. (引述理由解釋挑出 動的某個特徵，從 動主
體的觀點看來，這個特徵可以證成該 動、使它是可理解的。) This
is often what we are looking for when we ask why a person did
something – we want to know why they thought their action was
choiceworthy or made sense to do.
fi
行
行
行
 10

▸ 2. 的論的解釋：
A second way in which we can rationally explain an action is by
showing how it is related to some further goal or end the agent
has (這個 動 & 動主體進 步的 的):

Teleological: “Mariko ate/is eating a banana peel in order to get

more ber in her diet.”
目
fi
行
行
一
目
 11

▸ Teleological explanations are characteristically neutral with

respect to the ways, if any, in which the agent viewed her action as
good or desirable. Rather, they elucidate the action in purely
means-end terms, revealing what the agent ultimately aimed to
accomplish in undertaking that action. ( 的的解釋不涉及 動主體
視他的 動為好的或可欲的， 是只 段- 的的語詞去揭 他採
取這個 動是為了達到什麼最終 的。)
行
行
而
目
用
手
目
目
行
示
 12

▸ 3. 理主義的解釋：
Third, we might explain the action by pointing to certain
psychological features of the agent, thereby revealing her
understanding of the world and the motivations behind her deed:
(指出 動主體的某些 理特徵，以揭 他對於世界的理解以及他
動背後的動機。)

Psychologistic: “Mariko ate/is eating a banana peel smoothie

because she wants to lose weight and believes that eating less
fruit and more ber will cause her to lose weight.”
心
行
fi
心
示
行
 13

▸ The psychologistic form of explanation highlights the

importance of what the agent had in mind. After all, a reason
cannot explain an action if the agent did not believe it or was not
motivated by it. It might be that blending a banana peel is the
best way to sharpen the blades of one’s blender, but if Mariko did
not know this, or did not care about blender maintenance, then
the blade-sharpening properties of her action do not explain it.
 14

▸ 4. 素樸的解釋：
Fourth, we might simply cite another action that the agent was
engaged in: (引述另 個 動主體正在做的事 (或過去進 中的)，去
解釋該 動。)

Na ve: “Mariko walked/is walking into the kitchen because she is

getting a snack.”
ï
行
一
行
行
 15

▸ We can call this form of action explanation “Na ve” to contrast it

with the psychologistic form, which we might view as more
“sophisticated.” Na ve action explanation recharacterizes the
action to be explained in terms of another action that is or was in
progress. It reveals that on this occasion, Mariko’s act of walking
into the kitchen is also an act of getting a snack.
ï
ï
 16

▸ 5. 上述四種解釋分別強調了 動的理性的可理解性的某 個 向 (an

aspect of the rational intelligibility of action)，這些 向之中的 個
可能被其它 向所蘊含。

The Teleological guise brings out the way in which rational actions
tend to be goal-directed, undertaken as a means to some further
end. We gain understanding of what the agent is doing when we
discover what that end is.
面
行
面
一
面
一
 17

▸ The Psychologistic guise brings out the fact that the end must be
something that the agent herself has in mind and is moved by, as
well as that the means-end relationship might only exist in her
mind (if she is mistaken about whether her action will help her
achieve her goal). We gain understanding of her action by seeing
it from her perspective.
 18

▸ The Reasons guise brings out the fact that the agent will normally
be motivated to pursue an end because she sees it as desirable,
good, or just something that it makes sense to do. It increases our
understanding by revealing why the agent has the end that she
does.

And the Na ve guise invites us to take a step back and view what
is going on as a smaller part of a larger whole.
ï
 19

3.2 REASONS FOR ACTION: MOTIVATING VS. NORMATIVE

▸ 1. 關於「 動主體 動時所根據的理由」 (“the reason for which the
agent acted”)：
First, this is a philosophical term of art, and is meant to be
compatible with the thought that we often act for multiple reasons
( 動主體基於多個理由 動). It is possible to donate to charity
both for the reason that people are in need of help and for the
reason that it will reduce your taxes.
行
行
行
而
行
 20

▸ Second, the phrase is meant to get at what some philosophers call

“motivating reasons”: the reasons that actually move us to act (動
機理由：實際上推論我們去 動的理由). This is contrasted with
“normative” or “justifying” reasons: considerations that really are
good reasons for action (規範理由/證成理由：採取 動的真正的好
理由).
行
行
 21

▸ 動機理由 vs. 規範理由/證成理由：

This distinction is drawn because normative reasons can
sometimes fail to motivate us, while motivating reasons can
sometimes fail to be good reasons.
(例 ) In the above example, we can imagine that Mariko has a
good normative reason to eat the banana because bananas are a
source of needed potassium (鉀). She is not motivated by this
consideration, however, either because she does not know that
bananas are a good source of potassium or because she cares
more about avoiding calories than getting the nutrients she
needs. On the other hand, her motivating reason for eating the
peel might not be a good normative reason, if she does not in fact
need more ber in her diet.
子
fi
 22

▸ 2. 區分了動機理由 vs. 規範理由/證成理由，是否表 這是兩個類別的

理由 (two different categories of reasons)？

… some philosophers think that there are. But this may not be the
best way to conceive of the terrain here. After all, we should try to
avoid the rather depressing conclusion that we are never
motivated by good normative reasons.

示
 23

▸ (作者 (其實是 Dancy) 的看法： 者都是 動主體的考量，只是從不

同 向去看待他如何考量。) We can do this by holding that
motivating reasons and normative reasons are the same kind of
thing – that they are both “considerations,” say. When we call a
consideration a motivating reason, we are focusing our attention
on what mattered from the agent’s perspective, whereas when we
call it a normative reason, we are focusing our attention on
whether the agent was right to view things in this way.

▸ 動哲學主要關 的是理由作為動機理由。
At any rate, the philosophy of action is primarily concerned with
reasons in their motivating guise.
行
面
心
二
行
 24

▸ 3. 關於動機理由，有規定怎樣才能作為動機理由嗎？萬 它很奇怪
呢？
That said, there is philosophical disagreement over whether there
are normative constraints on what kinds of things can feature as
motivating reasons and thereby serve as rationalizing
explanations of action. Can just any consideration be the reason
for which the agent acted, regardless of how bizarre it is? Or must
motivating reasons bear suf cient resemblance to good
normative reasons if they are to explain action?
fi
一
 25

▸ (例 ) To illustrate the question, consider an example of

Anscombe’s in which a person claims to want a saucer of mud –
not because he wants to sculpt with it, or use it in a facial, or plant
a garden, but simply for its own sake. She asserts that we can
make no sense of this professed “desire,” and to that extent, we
might think that no mud-seeking action can be explained by it.
Other illustrations involve taking highly peculiar means to one’s
ends. If a person is motivated by thirst to put a dime in a pencil-
sharpener, or chooses to have coffee out of love for Sophocles (索
福克 斯，Oedipus 的作者), we might think that he fails altogether
to count as acting for a reason.
子
里
 26

▸ There are those who deny, however, that there are any such
inherent constraints on what we can desire or nd motivating
about an action. As long as the consideration did in fact motivate
the agent, no matter how bizarre, we can cite it as his reason in
explaining why he acted as he did.

fi
 27

3.3 MORE ON THE “WHY?” QUESTION

▸ 關於 the “Why?” 這個問題，它究竟在問什麼？我們可以從兩 來
考慮。

方
面
 28

▸ 1. 在蓄意 動的脈絡之外，當我們問「為什麼？」的時候，我們所尋
求的解釋經常是對於「事件 X 為什麼發 了？」(“Why did some
event X happen?”) 的答案。
(例 ) For instance, we might ask “Why did the plane crash?,”
seeking an answer like “Because the angle-of-attack sensors
malfunctioned.” (AoA，攻 ， 機的機翼弦線與 間 由流之夾
。)
這類解釋是在引述使得該事件發 的 個原因或條件，如果該原因或
條件未發 ，則該事件就不會發 (p.s. Lewis 的觀點)。
This kind of explanation tends to cite a cause or condition that
contributed to the fact that the event occurred, as well as to the
fact that it occurred when it did. Roughly, the implication is that if
the sensors had not malfunctioned when they did, the plane
would not have crashed on that particular occasion.
角
子
生
行
角
飛
生
生
一
生
飛
行
自
 29

▸ Among other things, this kind of understanding of past events

helps us to predict what will happen in the future (對於過去的事件
的這種理解 (i.e. 因果關係)，有助於預測未來的事件), such as that
there will be more crashes involving planes of this type (類 vs. 類).
 30

▸ 2. 在蓄意 動的脈絡之下，當我們問「為什麼？」的時候，(1) 我們
問的是：為什麼 個 動主體做了他所做的事 (why an agent acted
as she did)？我們想知道是什麼使得這個 動對他來說是合情理的、
值得做的 (what it was that made the action seem to the agent to
be sensible or worth doing)。
I.e. 尋找他的 (動機) 理由去合理化他的 動，使之成為可理解的。
行
一
行
行
行
 31

▸ 除此之外，(2) 我們也想問：這個 動為什麼發 了 (why the action

happened at all)？
(例 ) We want to know not only what Mariko was trying to
accomplish by putting the peel in the blender, but also why she
was making a smoothie at that time rather than studying or doing
nothing.
子
行
生
 32

▸ Thinking of the explanandum – the thing to be explained – in this

way has the appeal of locating human actions in a broadly
naturalistic order, in which they are on the same footing as other
events that we aspire to explain and predict. (以這種 式去思考被
解釋的 動，是把 類 動放在 個寬廣地 然的秩序中去了解，在
這個 然的秩序中， 類 動和其它我們想要解釋和預測的事件有著
相同的立 點。) If action explanations can tell us something about
why a particular action occurred, then they can also help us to
predict what people will do in the future (類 vs. 類).

▸ 不過，有些哲學家 (e.g. Jennifer Hornsby) 認為，在 動解釋中 (2)

(i.e. explaining why something happened) 是不必要的。
自
行
足
人
人
行
行
一
自
行
方
 33

3.4 ACTION EXPLANATION: FOUR VIEWS

▸ 那麼，合理化解釋 (rationalizing explanations) 是如何解釋 動的？
How does it work?

行
 34

▸ 3.4a 觀點 ：The Rational Interpretation View

▸ 1. 受到 Ryle 和 Wittgenstein 的啟發，合理化的詮釋的觀點主張，

動解釋的基本形式是理由 (reasons)。我們不 再找更根本的東 去
說明為什麼理解可以解釋 動，因為理性的活動是 個本 就有解釋
的模式 (rational activity is a pattern that is explanatory in its own
right)，這個模式關係到了何謂「理性的 (rational)」、「理性的 物
(rational creatures)」，後者的定義涉及了信念、欲望、和 動三者
之間的邏輯關係。

“explanatory”: giving an explanation about something (解釋的，說

明的)
力
一
行
用
一
身
行
西
生
行
 35

▸ No further account in terms of some deeper phenomenon is

required to understand how reasons explain what rational
agents do. Rather, the thought is that rational activity is [a
pattern] that is explanatory in its own right. How does this
work? Well, rational creatures by de nition tend [to desire what
they believe to be good, to hold beliefs that are supported by the
information they have, and to act in pursuit of what they desire
given what they believe].
fi
 36

▸ These observations are meant to be conceptual rather than

empirical in nature – it is simply what we mean by “rational.”
When we ask why an agent is doing something, according to this
view, we aim to interpret that particular action within this broader
pattern and thereby reach a deeper understanding or
“Verstehen.” We do this by redescribing the action in a way that
reveals the considerations in light of which it was intelligible from
the agent’s perspective.
 37

▸ (根據定義，理性的 會欲求他認為是好的的東 ，持有被他擁有的

訊息所 持的信念，並且，給定這樣的信念，他會採取 動去滿
他的欲求。當我們試著去理解 個 為什麼做了某件事，只要把他的
動放在這樣的 個 pattern 之中，透過他的信念、欲望重新描述他
的 動，我們就理解了他的 為，彷彿站在他的觀點去看事情。）
行
行
支
一
人
行
一
人
西
行
行
足
 38

▸ 2. (例 ) (為什麼 G 在深夜開 離開？) Suppose we see Gilbert

driving off in his car late at night and wonder why he would do
such a thing. The rst step in interpreting his action is to use the
context, behavioral evidence, and Gilbert’s own reports in order
to redescribe the deed in terms of a more general purpose: he is
going to the store (因為他要去商店。).
(為什麼他要去商店？) We then attempt to locate what kinds of
motivations and beliefs could have featured as the premises in
Gilbert’s reasoning that concluded in going to the store. If we
learn that he wanted some chocolate, we can infer that he
believed he could buy some at the store (因為他想要來些巧克 ，
且他相信商店可以買到巧克 。).
而
子
fi
車
力
力
 39

▸ 所以，為什麼給了 動主體的理由就能解釋他的 為？因為 個理性

的 就是會依據他的理由去 動。
We gain understanding of what he is doing because this is just the
kind of thing a rational agent (理性的 動主體) does when he
wants some chocolate, has a car available, and has nothing else
better to do. His action is explained by showing that it has the
structure that is characteristic of rational activity (理性的活動). Call
this the “Rational Interpretation” view of how action explanations
work.
人
行
行
行
行
一
 40

▸ 3. 但是， 並是完全理性的，有時候我們會欲求我們不應該欲求的東
(可能它不是好的)，相信不應該相信的事物 (可能缺乏證據) 等。合
理化的詮釋的觀點有辦法解釋以這類理由為依據的 動嗎？

Still, many of these rational aws follow predictable patterns. We

tend to prioritize short-term pleasures like smoking and tanning
over our long-term health, while our beliefs are shaped by wishful
thinking, con rmation bias (確認偏差), and base-rate neglect (忽略
基本比率).
Thus, the understanding provided by a reasons explanation need
not depend on ideal rationality; the pattern can be stretched to
accommodate common foibles.
西
人
fi
fl
行
 41

▸ 有什麼 動是合理化的詮釋的觀點解釋不了的嗎？
What it cannot allow for is the extreme irrationality of seeking to
acquire mud for no further purpose or putting a dime in the
sharpener because one wants a soda.
行
 42

▸ 4. 是什麼使得 個合理化的解釋為真或為假？
But what makes a candidate rational explanation either true or
false? As long as we can make good rational sense of Gilbert by
attributing a desire for chocolate and a belief that there is
chocolate to be had at the store, can there be any further question
about whether this explanation is the right one?
一
 43

▸ 第 稱權威？
Those who favor this approach take the kind of understanding we
seek to be primarily rst-personal, and so will prioritize the agent’s
own report of her reasons for action. Indeed, a central idea found
in the work of both Wittgenstein and Anscombe is that the agent’s
answer is not really a report of some independent fact, but rather
an expressive act that has the power to make the connection
between reason and action and thereby determine the fact of the
matter. In other words, the reasons for which we act are
normally what we sincerely assert they are.
一
人
fi
 44

▸ 為證據？
This rst-personal authority is not absolute; it can be defeated by
strong behavioral evidence that con icts with the agent’s own
pronouncement. If we know that Gilbert doesn’t have a sweet
tooth and that he often nds excuses to visit the store when a
certain cashier is on duty, we might well disregard his claim that
he is motivated by chocolate. This kind of case must be the
exception, however, in the sense that for Gilbert to be a rational
agent at all, the answers he offers to the “Why?” question must
normally be accepted as decisive.
行
fi
fi
fl
 45

▸ 5. 合理化的詮釋的觀點認為， 動的理由 ≠ 動的原因。

(主張) Most importantly, the Rational Interpretation view denies
that rationalizing explanations have anything to do with
identifying the causes of the agent’s behavior. (動機) In this, it
is motivated by the desire to avoid “scientism (科學主義)”: an
excessive, even slavish tendency to think that science is the only
genuine source of knowledge about reality.
行
行
 46

▸ (論證) Proponents of the view have argued that because (1， 來

理解 動的那個理性的 pattern 在本質上是邏輯的) the rational
patterns picked out by action explanations are logical in
nature, the agent’s desires and beliefs could not be connected
with her actions as cause to effect. Roughly, the thought is that
“action,” “belief,” and “desire” are conceptually inter-de ned
(「 動」、「信念」、和「欲望」是概念上相互定義的): desires
just are states that produce action, given our beliefs; beliefs just
are states that guide action, given what we desire; and action just
is the product of wanting something and believing we can get it
( 動是 組欲望和信念的產物).
行
行
行
一
fi
用
 47

▸ David Hume in uentially claimed, however, that (2，原因和結果必

須是可區別的存在) cause and effect must be “distinct
existences”: among other things, effects cannot be logically
inferred from their causes (結果不能是從原因邏輯地被推論出來
的). Otherwise we would be able to know about causal
relationships in the world without any kind of empirical
investigation, and this is not in fact possible.
fl
 48

▸ Advocates of the “logical connection” argument conclude that (所

以，3，兩個東 之間有邏輯關係，它們之間就不會有因果關係，那
麼，信念、欲望、和意圖不能是 動的原因) entities standing in
logical relations to one another cannot also stand in causal
relations, and that beliefs, desires, and intentions therefore could
not be causes of action.
西
行
 49

▸ As A.I. Melden puts the claim, “. . . where we are concerned with

explanations of human action, there causal factors and causal laws
in the sense in which, for example, these terms are employed in
the biological sciences are wholly irrelevant to the understanding
we seek” (1961, 184).

What [is denied] is that desires serve as causes of action when

they are cited as a way of picking out the agent’s reasons.
 50

▸ 3.4b 觀點 ：The Causal Theory of Action Explanation

▸ 6. 合理化的詮釋的觀點的 動解釋並不是科學所給出來的那類解釋，
這使它遭遇了 些困難。
... how the very same occurrence can be given more than one
complete explanation. After all, actions do seem to be the kind of
thing that we can study using the frameworks of biology,
psychology, and neuroscience. If we can give a complete
explanation of why some behavior occurred from a scienti c point
of view, this threatens to render the contribution of a rational
interpretation mysterious, if not entirely super uous.
二
一
行
fl
fi
 51

▸ 相對地， 動解釋的因果理論則想要提供 個與經驗科學更 致的解

釋。
The Causal Theory of Action Explanation aspires to offer an
account that is potentially more uni ed with the empirical
sciences.
行
fi
一
一
 52

▸ 7. The view is standardly credited to Davidson and his seminal

1963 paper “Actions, Reasons, Causes,” although it actually has a
long history that traces back at least to Aristotle.

合理化的詮釋的觀點反對把 動的理由視為 動的原因，我們已經知

道他們的論證之 ，i.e. the logical connection argument：
… because the concepts “desire,” “belief,” and “action” are
logically related to one another, desires and beliefs could not be
causes of action, for causes cannot logically entail their effects.
一
行
行
 53

▸ 但是，Davidson 反對這個論證，因為它混淆了關於概念的事實和關
於本體論的事實：因果關係是個別的事件之間的形上的關係，這個關
係是否成立，與我們如何描述這些事件無關， 於兩個事件之間是否
有邏輯關係，是看我們使 什麼概念、如何描述它們。
… this argument is actually invalid. The mistake arises from
confusing facts about concepts (the ways in which we human
beings think about the world) with facts about ontology (the way
the world is independently of how we conceptualize it). On a
widespread way of thinking about it, causation is a metaphysical
relation that holds between particular events no matter how we
describe them, whereas logical relations hold only between
concepts or descriptions.
用
至
 54

▸ (例 ) For example, suppose that the loud thunderclap caused my

cat Blinky to startle.
This will remain true even if we change Blinky’s name to Ringo, or
if we describe the events as “a sudden noise” and “the black cat
startled” – the descriptions do not matter, as long as they succeed
in picking out the same events.
We could even choose descriptions that are logically related to
one another; for instance, we could refer to the thunderclap as
“the cause of Blinky startling” and assert the trivial logical truth
that “The cause of Blinky startling caused Blinky to startle.” But the
fact that the two events are logically related under these
particular descriptions does not show that they are not also
causally related qua events. (兩個事件在某個描述之下邏輯地關連
在 起，不表 這兩個事件不能也因果地關連在 起。)
一
子
示
一
 55

▸ Thus, the fact that action explanations serve to render actions

intelligible in a way that exploits certain conceptual interrelations
does not rule out the possibility that there is also a causal
relationship at work between the events we pick out by using
these concepts in an explanatory context. ( 動解釋利 某種概念
的相互關係使得 動是可理解的 (i.e. 引述 動的理由來合理化該
動)，不能排除被這些概念所描述的事件之間有因果關係的可能性。)
行
行
行
用
行
 56

▸ 假設「A 引起 B」為真，那麼 B 的原因即是 A，我們可以把該語句改

寫為「B 的原因引起 B」，這個新的語句是分析的 (Davidson 2001:
14)。
(例 )「敲鐘 (the striking of the bell) 引起鐘響 (the bell’s
tolling)」，敲鐘是鐘響的原因，「鐘響的原因」和「敲鐘」是共同
指涉詞，現在我們 「鐘響的原因」來替換「敲鐘」，得到「鐘響的
原因引起鐘響」這個因果陳述句，它是 個分析地、先驗地為真的語
句 (Mele 2003: 68)。
由此可 ， 個因果陳述句的真，依賴於被描述的兩個事件的確有因
果關係， 於 個因果陳述句究竟是分析的或是綜合的，端看事件如
何被描述。
子
見
至
一
一
用
一
 57

▸ 8. 動解釋的因果理論訴諸 理狀態去解釋 動。
According to the Causal Theory of Action Explanation, the most
fundamental form of rationalizing explanation is Psychologistic (合
理化解釋的最基本的形式是 理的). The basic idea is that actions
are explained in part by citing the psychological activity that
caused them to occur ( 個 動可以透過引述引起該 動的 理活
動 被解釋). According to Davidson’s version of the view, the
relevant mental causes are a combination of something the agent
believes and something she desires, or otherwise has a “pro-
attitude” about (能解釋 個 動的 理的原因，是由 個信念和
個欲望 (或 個正 的態度) 所構成的 個組合). “Pro-attitude” is a
term of art Davidson uses to refer to a very wide variety of ways in
which one can be attracted to a state of affairs: one can desire it,
judge it to be good, right, or beautiful, simply feel a yen toward it,
and so on.
而
行
一
面
一
一
心
心
行
行
心
一
行
一
行
心
一
 58

▸ (例 ) In our earlier example, Gilbert has a pro-attitude toward

eating chocolate and believes that he can acquire some at the
store. Roughly, the claim is that together, these two mental states
(sometimes called a “belief–desire pair”) play a causal role in
bringing it about that he drives to the store. Other versions of the
Causal Theory may include additional mental states like intention
as potential causal antecedents of action.
子
 59

▸ 動的 理解釋既是 種因果解釋 (i.e. 訴諸 理狀態作為引起 動

的原因)，也是 種合理化的解釋 (i.e. 這些 理狀態使得 動是可理
解的)。
It is crucial to emphasize that in saying that psychologistic
explanations of action are a species of causal explanation, we are
not denying that they are also rationalizing. To succeed, the
explanation must also justify what was done, or at least render it
somewhat rationally intelligible.

In this respect, the view aims to incorporate the insights of the

Rational Interpretation model while at the same time going
beyond it.
行
心
一
一
心
心
行
行
 60

▸ 9. 動解釋的因果理論主張合理化的解釋亦是 種因果解釋，I.e. 合
理化解釋不僅是提到某些 持採取 個 動的理由 (some
consideration that favored performing the action)， 是提供了
動主體據以 動的理由 (the reason for which the agent acted)。(後
者即是這個理論比合理化的詮釋的觀點更進 步的主張。) 為什麼要
這麼主張？
行
行
支
一
行
一
一
而
行
 61

▸ Davidson 的 標是要解釋這兩者之間的區別：有理由 (去做某事)

(having a reason) (to do something) vs. 為了那個理由 動
(acting for that reason)

Davidson’s Challenge: a successful theory of action explanation

must explain what is added to the fact that S had reason R to
perform action A when we say “S-ed because R.”
(Davidson 的挑戰： 個成功的 動解釋的理論，必須解釋：當我們
說 “S A-ed because R”，什麼東 被新增 S 有理由 R 去採取 動 A
這件事？)
目
一
行
西
至
而
行
行
 62

▸ (例 ) … of a case in which there are multiple reasons in light of

which the agent could have acted, but where he in fact acted for
only one of those reasons. Suppose Kunal’s great-aunt has left him
her fortune in her will and has made clear to him that she does not
wish the doctors to take extraordinary measures to prolong her
life. When she falls ill, Kunal has at least two reasons to prevent the
doctors from resuscitating her: it will hasten his acquisition of her
fortune, and it will respect her wishes. …
子
 63

▸ 看起來，兩個理由 樣可以解釋 A，我們無法知道到底哪 個才是正

確的解釋。
… the data available for interpretation may include the whole
history of the episode, minor tics in his behavior, and so forth. But
this does not ensure that there will be only one way to rationalize
Kunal’s deed, and insofar as there are multiple interpretations that
t equally well, they are all equally good as interpretations.
Nothing further can be said about whether Kunal really acted
because he wanted the money or because he respected his great-
aunt’s wishes, since attributing either of these reasons would
make good rational sense.
fi
一
一
 64

▸ 僅當 動的理由是 動的原因，我們才能挑出正確的解釋。
A reason the agent had, but didn’t act for, is a belief–desire pair
that did not in fact play a causal role in bringing her action about.
And if she acted for a reason, then there must have been a
relevant belief–desire pair that did play a causal role. If we
seek to understand the real reason why Kunal prevented his great-
aunt from being resuscitated, we must inquire into what actually
caused his behavior.

(先前的例 ) … But it might be that he does it only out of a desire

for her fortune, even though he also desires to respect her
wishes.
行
子
行
 65

▸ … explanations of action mentioning the agent’s beliefs and

desires are not only rationalizing, but also a species of causal
explanation.
 66

▸ 10. 接下來是關於 動解釋的因果理論的 些澄清。

先，這個理論是要回答「理由如何解釋 動？」(“How do reasons

explain action?” ) 它可能不必然是要去回答「是什麼使得 個 動是
願的或蓄意的？」這個問題 (“What makes an action voluntary or
intentional?”)。為第 個問題提供回答的是另 個理論，i.e. 動的
因果理論 (the causal theory of action)，它主張 個 動發 在 個
正常的因果鍊裡，是使得它是 個蓄意 動的充分必要條件。(...
having the right causal history is a necessary condition of being
an intentional action? [Or,] … being caused in the right way is
both necessary and suf cient for being an intentional action?)
首
自
行
二
fi
一
行
一
行
一
一
行
一
生
行
行
一
 67

▸ We will examine this ambitiously reductive view about the

metaphysics of action in Chapter 4. The two views often get
blurred together under the label “causalist,” so it is important to
note that they are in fact distinct.
 68

▸ 第 ，在討論主體性的語境中，我們可以區分出兩種不同的因果關係
(causation)。
其中 種對於因果關係的看法是 Aristotle 所謂的「動 因」
(ef cient cause)，改變或靜 的 要來源 (the primary source of
change or rest)。
(例 ) This is the kind of causation at issue in claims like “the cue
ball caused the eight ball to roll into the pocket” and “the
assassination of Archduke Ferdinand caused World War I to break
out.”
二
fi
子
一
止
首
力
 69

▸ 這也是 然科學中所使 的「原因」和「結果」 的概念，因果關係

是在於兩個事件之間，其中 個作為原因、另 個作為結果。
On the standard way of thinking about ef cient causation, the
causal relation holds between two events: one event that is the
cause, and a second event that is the effect.
(例 : 不是 球這個物體作為原本， 是帶有動 的 球撞擊八號球
這個事件作為原因。) On this view, it is not strictly speaking the cue
ball that causes anything, but the event of the cue ball striking the
eight ball with a certain momentum.
子
自
白
用
一
而
fi
一
力
白
 70

▸ 有些 認為因果關係在於兩個事件之間、或在於 個情況中的兩個特
徵之間。
There are also variations of the view that accord a less
fundamental role to events, allowing the causal relation to hold
between facts or other features of a situation.
人
一
 71

▸ 另 種對於因果關係的看法是 “agent-causation” ( 動主體因果關

係？)，當 動主體 動時，把 動主體本 視為原因。
… there is a distinct kind of “agent-causation” that is at issue when
someone acts. The agent-causal relation does not hold between
two events or facts; rather, the cause is held to be the agent
herself, and not merely some event that she participates in or
some proper part of her.
(例 ) In other words, if the cue ball were an agent, then it would
be accurate to say that the cue ball itself was the fundamental
cause of the eight ball’s rolling into the pocket – not the
movement of the cue ball, or the event of its striking the eight
ball, but simply the ball, viewed as a substance.
一
子
行
行
行
身
行
 72

▸ This is not an understanding of causation that is familiar from

science, but the intuition is that an action must be caused by
nothing less than the agent as a whole (stemming, often, from a
conception of action that prioritizes attributability).

We will brie y return to the agent-causal view in Chapter 4, but to

be clear, the notion of causation at issue in discussing Davidson’s
view and its descendants is ef cient causation and not agent-
causation.
fl
fi
 73

▸ 第三， 動解釋的因果理論有時候以「理由即是原因」(“reasons are

causes”) 這個標語被表述，但這是有誤導性的，特別是在考慮規範理
由和動機理由的區分之下。

The view holds that the causes of action are mental states like
beliefs and desires – or more precisely, mental events that are the
manifestations of such states. If “reasons are causes,” then reasons
for action would turn out in every case to be belief-desire pairs.
( 動的原因是信念、欲望等 理狀態，如果理由即是原因，那麼
動的理由是信念、欲望等 理狀態。)

But this is a highly unappealing implication, whether we mean

“normative reasons” or “motivating reasons.”
行
行
心
心
行
 74

▸ 假如「理由即是原因」意思是「規範理由即是原因」(“normative
reasons are causes”)，這表 持採取某個 動的理由
(considerations that count in favor of doing anything) 是關於我們
的 理狀態的某些事實 (facts about our own mental states)。但
是，關於我們的 理狀態的事實鮮少為 動提供規範理由，規範理由
通常被認為是非 理的事實 (non-mental facts)。
(例 ) As Jonathan Dancy points out, there are rare cases in which
facts about our own beliefs and desires do provide normative
reasons for action, as when the fact that you believe the CIA is
spying on you, or the fact that you desire to hide under your
covers all day, are reasons to consult a psychiatrist. These cases
are very different, however, from the normal case in which our
reasons are non-mental facts. It is the fact that ber is good for the
digestion, together with the fact that banana peels contain ber,
that gives Mariko a reason to eat a banana peel.
心
子
心
心
示
支
行
行
fi
fi
 75

▸ 假如「理由即是原因」意思是「動機理由即是原因」(“motivating
reasons are causes”)，似乎是比較合理的。
After all, the main impetus for introducing the category of
motivating reasons is to capture the agent’s perspective on her
action, whether or not that perspective is mistaken. Further, the
notion of “motivation” and the notion of “cause” seem to be
connected.
 76

▸ 這樣我們可不可以說，所有的動機理由都是 理狀態、規範理由都是
事實？
Why not then say that all motivating reasons are mental states,
whereas normative reasons are worldly facts?
But as noted in Section 3.2, dividing these two categories so
sharply would make it impossible, or at least rare, to act for good
normative reasons. The considerations that justify our actions
would almost never be the reasons that actually motivate us. (可能
的蘊含：我們幾乎不會為了規範性的理由、好的理由去 動？) This
is also an unhappy result.
心
行
 77

▸ 可能的解套 式：動機理由不是命題態度， 是命題態度的內容。當

信念的內容為真、欲望的內容確實是可欲的，那麼動機理由也可以是
規範理由。反之則不然。
One plausible solution is to hold that motivating reasons are the
contents of our attitudes: not our beliefs and desires, but what we
believe, in combination with what we desire.
(例 ) What rationalizes Mariko’s smoothie-drinking is the ber in
the peel, as represented by her beliefs, and the advantages of
good digestion, as represented by her desires.
When our beliefs are true, and when we desire what is genuinely
desirable, then our motivating reasons are also good normative
reasons. And when our attitudes are out of kilter with reality, our
motivating reasons are not good normative reasons.
子
方
而
fi
 78

▸ 第四，信念、欲望這些 理狀態能作為 體運動這樣的物理事件的原

因嗎？
A nal complication concerns the topic of mental causation more
generally. Is it plausible to suppose that mental states like beliefs
and desires are ever the true causes of physical events like bodily
movements?
fi
心
身
 79

▸ 如果把信念、欲望視為 腦的神經狀態，它們或許就有資格作為 體
運動的原因 (i.e. 物理事件或狀態引起物理事件)。但問題是，如此
來，我們要 神經的、物理的語詞去描述信念、欲望， 不是使 通
俗 理學 (常識 理學) 的語詞去描述信念、欲望。
Assuming that (in human beings, at least) the mind is realized by
the brain, the particular states that we pick out using mental
concepts like “belief” and “desire” are neural states. It is thus
tempting to think that the causal story of how our bodily
movements are initiated and guided must ultimately be couched
in terms of neural events and other physical states of the
organism, rather than in the vocabulary of “folk psychology.”
心
用
心
大
而
身
用
一
 80

▸ It is still very much an ongoing debate in the philosophy of mind

whether or not familiar mental concepts or properties will feature
in a mature account of the causes of human behavior.
SEP 81

▸ 什麼是 folk psychology?

(https://plato.stanford.edu/archives/fall2021/entries/folkpsych-
theory/)

Folk psychology is a name traditionally used to denote our

everyday way of understanding, or rationalizing, intentional
actions in mentalistic terms. This quotidian competence is known
by other names in the philosophical literature: commonsense
psychology; naïve psychology; Homo sapiens psychology; the
person theory of humans; the intentional stance; propositional
attitude psychology; belief-desire psychology (see, e.g.,
Churchland 1979; Dennett 1987; Goldman 2006; Bogdan 2009).
SEP 82

▸ There is great interest in folk psychology not only because of its

status as a familiar way of making sense of our actions and those
of others but also because it is thought to underwrite a range of
moral, legal, educational, clinical, and therapeutic practices
(Fodor 1987; Baker 1988). For this reason, a great deal of work in
analytic philosophy has been devoted to better understanding
folk psychology and its cognitive basis.
 83

▸ Davidson 怎麼看待這個問題？

It may seem that the Causal Theory of Action Explanation

prejudges this question by insisting that beliefs, desires, and other
pro-attitudes are causes. This is not necessarily so.
 84

▸ Davidson himself denies that mental concepts gure in causal

laws, and takes mention of mental states like belief and desire to
be mere shorthand for some associated event that was the
catalyst of the relevant behavior. The Causal Theorist is committed
only to thinking that when we give rationalizing explanations of
action, there is a causal connection in the vicinity that partially
grounds the truth of the explanation, even if that connection can
only be formulated precisely using non-mental concepts. In this
respect, it is not necessarily a reductive view about the
relationship between the mental and the physical (though any
given version might well be). fi
 85

▸ 3.4c 觀點三：Teleological Realism

▸ 11. 在解釋 動上， 理的形式 (psychologistic form) 被因果理論認

為是 要的， 的實在論則把 的的 (Teleological) 視為是 要的。

的實在論對 動的解釋具有下列形式：
Teleological explanations, recall, have the form “S A-ed in order to
B” or “S’s A-ing was directed at B.”
目
首
行
目
行
心
目
首
 86

▸ (因果理論將訴諸 理狀態或事實去使 該形式。) According to the

Causal Theory, when claims like this are true, it is because of some
underlying fact about the agent’s psychology – that S wanted to
bring about B, or believed that A-ing was a way of bringing about
B.
( 的實在論反對 的的事實可以被化約為任何非 的的現象。) In
contrast, the view we will call Teleological Realism denies that
such teleological facts are reducible to some other, non-
teleological phenomenon. Rather, it accepts that teleology is part
of the basic furniture of the world.
目
心
目
用
目
 87

▸ 12. 的的實在論：
蓄意 動是有意識地導向某個 的的 動， 動解釋中的「因為」是
要去揭露那個 的是什麼。
The Teleological Realist view has been most prominently
defended by George Wilson and Scott Sehon. The basic idea is
that intentional action is behavior that is “sentiently directed”
at some goal, and that the “because” of action explanation
functions to reveal what that goal was.
目
行
目
目
行
行
 88

▸ 問 個 他為什麼做了某件事，是要去誘出他的 動所導向的更進
步的 的。回答這個問題最直接的 式，是去說「他為了 B 做了
A。」
When we ask why Kunal prevented the doctors from resuscitating
his great-aunt, we are aiming to elicit the further end at which his
action was directed. The most straightforward way of answering
the question will be to say that he did it in order to inherit the
money, …
一
目
人
方
行
而
一
 89

▸ 訴諸信念、欲望等 理狀態去解釋 動，只不過是 間接的 式去說

明相關的 的 已。
… but the other guises of explanation can be understood as
oblique ways of specifying the relevant goal. Speaking of what
Kunal desired and believed, for instance, is simply a different way
of illuminating the state of affairs his behavior was directed at
achieving.
目
而
心
行
用
方
 90

▸ 13. 對 的的實在論者來說，「導向」(“directing at”) 不是 個因果

的概念，他們不是要給 動 個因果的解釋。 個 動的原因，應該
是 神經 理學的語詞去表述。
Importantly, though, “directing at” is not meant to be an ef cient-
causal notion; the answer to the question of what Kunal’s behavior
was directed at is not also an explanation of what caused his
mouth to open and utter the words “do not resuscitate.” The latter
question will, presumably, be answered in neurophysiological
terms.
用
目
生
行
一
一
行
一
fi
 91

▸ 為什麼 的的實在論者不認為 個 動的 的是該 動的原因？

，他們懷疑通俗 理學的科學的有效性。
Skepticism about the scienti c validity of folk psychology, which
understands the mind in terms of attitudes like desire, belief, and
intention, is a major motivation for this approach. If a completed
science of the mind and related phenomena such as motor
control (運動控制) will not involve anything recognizable as beliefs
and desires, then we might well think that any causal explanations
invoking such items will turn out to be mistaken. ... the
Teleological Realist proposes that we understand them in non-
causal terms. If their explanatory power derives from teleology
rather than causation, then they need not be in competition with
scienti c explanations.
一
方
面
fi
目
心
fi
一
行
目
行
 92

▸ 另 ， 動解釋的因果理論遭遇了因果的異常的難題 (the
problem of causal deviance)，這使得 的實在論者懷疑它真的能解
釋 動。
A second motivation for the view stems from skepticism that
causation could provide the illumination that we seek from action
explanations. Teleological Realists like Wilson and Sehon have
charged the Causal Theory with suffering from a fatal aw known
as the problem of causal deviance.
一
行
方
面
行
目
fl
 93

▸ 因果的異常的難題：
In essence, the worry is that being caused and rationalized by a
belief-desire pair is not enough for an event to have the status of
an intentional action, if the event is caused in the wrong way.
(例 ) In Davidson’s own classic example, a mountain climber has
lost his hold and a second climber is holding him on a rope, at
great peril to his own life:
A climber might want to rid himself of the weight and danger of
holding another man on a rope, and he might know that by
loosening his hold on the rope he could rid himself of the weight
and danger. This belief might so unnerve him as to cause him to
loosen his hold, and yet it might be the case that he never chose
to loosen his hold, nor did he do it intentionally. (1971, 79)
子
 94

▸ 這個例 如何對因果理論造成威脅？(其實主要是對 Causal Theory

of Action 造成威脅，詳 下章。)

The point is that the causal chain leading from his belief and
desire to his letting go of the rope is “deviant,” relative to the way
the Causal Theorist thinks actions are normally caused by our
beliefs and desires. …
子
見
 95

▸ … On the face of it, [the Causal Theory of Action Explanation] is

committed only to the following claim: if an event is the kind of
thing that can properly be given a rationalizing explanation, then
the truth of the explanation depends partly on the right kind of
causal connection holding between the agent’s attitudes and her
behavior. In cases of causal deviance, the Causal Theorist may
simply say that the resulting event is not subject to a rationalizing
explanation – after all, the example stipulates that it happened
because the climber was nervous and lost control. The view is
simply not committed to counting everything that is both caused
and rationalized by the agent’s beliefs and desires as an
intentional action.
 96

▸ Still, the possibility of causal deviance does put pressure on the

idea that causation plays any signi cant role in rationalizing
explanation. It turns out that causation alone is not enough to
meet Davidson’s Challenge of explaining the difference between
having a reason and acting for that reason. …
fi