Download Chapter 12 - What Is Knowledge?

Chapter Twelve
Descriptions, Prescriptions and
the Limits of Knowledge
Abstract: The concept of a ‘proposition’ is discussed. It is concluded that (1) stipulative
definitions, (2) the axioms and inference rules of formal deductive and metaphysical
systems, (3) 'sufficient evidence' assertions, and (4) ethical assertions are prescriptions
and cannot replace p in 'S knows p.' In contrast, empirical statements, true-in-a-language
(e.g. analytic) sentences, applied mathematical/deductive entailments, reportive
definitions, theoretic definitions, aesthetic assertions and social science theories are
descriptions, and potentially true and knowable. The Frege-Geach problem is addressed.
Metaethical prescriptivism is compared to expressivism. Prescriptions are compared to
performative utterances. The descriptive-prescriptive distinction is compared to the
declarative-interrogative-imperative distinction.
In this final chapter I summarize the limits of human knowledge. The definitions
of 'description' and 'prescription' are repeated below:
A 'description' is an assertion that purports to express a correspondence (or a
reflection) of some state of affairs, where its correctness (or incorrectness) is
independent of its acceptance (or non-acceptance) by particular persons.
A 'prescription' is an assertion that purports to express a stipulation (or rule)
upon a practice, where its correctness (or incorrectness) is dependent upon its
acceptance (or non-acceptance) by particular persons.
These definitions are intended to be objectively true theoretic definitions. A theoretic
definition is correct (i.e. true) only if its definiens accurately describes the phenomena
(object) being defined. The descriptive-prescriptive distinction represents an objective
feature of the world about what persons can know and mean.
Two of the major
conclusions in this book are that empirical assertions are knowable (chapters one through
four), while moral assertions are not knowable (chapter five). 1
1
To briefly summarize the first five chapters: Chapter one theorizes a predominately externalist (PE)
definition of 'knowledge' that states the conditions for when knowledge is obtained (and when it is not
obtained). Chapter two addresses radical skepticism and (the fact) I cannot know that I-am-not-adisembodied-brain-in-a-chemical-vat. Chapter three solves the skeptical regress-of-reasons problem,
claiming that assertions of what constitutes 'sufficient evidence' to know p, terminates in a persondependent prescriptive mode, but that this fact doesn't harm our ability to possess the necessary and
sufficient conditions for knowledge. Chapter four addresses Hume's problem of induction, and concludes
that persons can be 'personally justified' in believing that 'the past will resemble the future,' but cannot have
a truth-connecting 'justified belief' about this proposition because there are no non-circular arguments to
defend this belief. Chapter five introduces prescriptivism that maintains ethical assertions may be approved
(or disapproved) by humans, but that they are neither true nor false.
-2In contrast to the objective definitions of 'description' and 'prescription,' I suggest
that the following definitions of 'objectivity' are stipulations being consistent with the
descriptive-prescriptive distinction, the correspondence theory of truth, and ordinary
linguistic usage and belief:
A description is objectively true if it expresses a correspondence (or a reflection)
to some state of affairs that is independent of its acceptance (or acknowledgment)
by particular persons.
A description is objectively false if it doesn't correspond to, or reflect a state of
affairs.
These definitions are the best characterizations of our normal conception of 'objective'
truth and falsity.
With a summary of what descriptive propositions are, and what
prescriptive propositions are, we should say something about what a ‘proposition’ is.
The term 'proposition' has a broad use in philosophy. It is most often associated
with the concepts of sentence and truth. Unfortunately, there is no standard definition of
what a proposition is. What is a proposition? Some philosophers claim that it is 'the
object of an attitude.' We won't evaluate that claim here. 2
In its most neutral
characterization, a 'proposition' is a complete sentence asserted in a context that presents
the contents of one's thought. The pragmatic need for a distinction between 'sentence'
and 'proposition' is summarized with the following standard (and non-standard)
intuitions:
a) The sentence 'It is now raining' as a linguistic expression is not by itself
literally true or false. The sentence needs to be asserted in an environment and at
a certain time to be true or false. It is the proposition expressed (in a context) by
the sentence 'It is now raining' that is true when it is raining, and false when it is
not raining. Sentences are not literally true or false, but it is their utterance as a
'proposition' in a context that is either true or false.
b) The English sentence 'Snow is white' expresses the same proposition as the
German sentence 'Der Schnee ist weis.' Given that these sentences are different,
it isn't the linguistic entities (i.e. sentences of different language) that make the
utterances true, it is the proposition (that is their meaningful content) that is true.
2
It is claimed by some philosophers that a proposition is the object of a 'propositional attitude' as a relation
between a person and a proposition p. A person can have attitudes e.g. believe, doubt, hope, wish, expect,
desire, fear, or remember p and so on. Propositions can also be the referent of that-clauses. This
supplementary view about propositions cannot be discussed here.
-3c) The sentences 'Here is the red book' and the 'The red book is here' when
asserted in a context express the same (true or false) proposition. Since different
sentences are used, it is not the sentence that is literally true or false, it is the
proposition expressed by a sentence that is true or false.
d) 'Mark Twain wrote Huckleberry Finn' and 'Samuel Clemens wrote
Huckleberry Finn' are different sentences. Since Mark Twain is Samuel
Clemens, one might expect that these sentences express the same proposition.
But whether these sentences express the same proposition rests upon a person's
background knowledge. If a person doesn't know in a context that the two proper
names designate the same person, these sentences will express two propositions.
e) The sentence 'Persons should not smoke tobacco' is understood by a
metaethical cognitivist as a true or false proposition. For a noncognitivist, this
can be interpreted as a prescriptive proposition.
In defining a 'proposition' as a sentence asserted at a time in a context that presents the
contents of one's thought, this concept is not defined as always being representational of
something external to the mind (and either true or false). Of course, this interpretation of
a proposition and the intuitions of (d) and (e) go against tradition and will be strongly
resisted by philosophers who maintain that all declarative propositions, assertions, and
statements are exclusively true or false.
With the above summary of what descriptions and prescriptions are, and what a
proposition is, we turn attention to cataloguing the kinds of assertions that are descriptive
or prescriptive from an epistemic point of view.
I. Stipulative Definitions Are Prescriptions
In chapter six, 'stipulative definitions' were identified as prescriptions
A 'stipulative definition' introduces a specialized definiens for a definiendum.
This occurs in one of the following three contexts: (a) the initial naming of an
entity where the entity is newly-discovered, newly-introduced, newly-created, or
newly-renamed, or (b) in the notational abbreviation of one linguistic expression
for another (meaningful) linguistic expression, or (c) in a precise formalization
where a reportive definiendum-to-definiens relation is generally affirmed but a
definiens alteration (or explication) is proposed for pragmatic, technical, or
personal reasons.
-4The tripartite thesis is a hypothesis about the actual modes of how persons can specify (or
assert) the use of a linguistic symbol (in both natural and artificial languages). In chapter
six, it was concluded that stipulative definitions are prescriptions, and not descriptions.
When discussing formal deductive models, some philosophers have said that a model's
stipulative definitions are true by convention.
This locution can lead to a serious
misconception. To believe that a stipulative definition can be either true or false is a
major epistemic error.
II. The Axioms and Definitions of Measurement Systems Are Prescriptions
In chapter seven it was argued that the primitive symbols, definitions, axioms,
grammar, and rules of inference, found in formal deductive systems are introduced as
prescriptions. In chapters seven, nine, and ten, important metaphysical and semantic
axioms (e.g. law of identity, law of bivalence, and compositionality) were interpreted as
prescriptions. Deductive systems are introduced with the pragmatic purpose of
measurement; e.g. physical space is measured in geometry, quantity in arithmetic, valid
arguments in deductive logic. Whether one measurement system is better than others
depends upon whether it helps us better understand or quantify the domain involved.
With respect to the pragmatic purpose of measurement, it should also be observed
that in the social sciences, for example, within the discipline of Economics, that semidescriptive assumptions (viz. stipulations) play a prominent role in theory construction.
In responding to the question in the title of his essay “Are General Equilibrium Theories
Explanatory?” (1984) Daniel M. Hausman presents eight ‘lawlike’ assertions behind
general equilibrium theories, where ‘economic equilibrium’ is defined as “a state of
affairs where there is no excess demand- a state of affairs in which at the going prices
nobody wants to go on exchanging.” From this fixed definiens concept, eight statements
of an ‘equilibrium theory’ are described as the basis of neo-classical economics (p.345):
(1) For any individual A and any two options x and y, one and only one of the
following is true: A prefers x to y, A prefers y to x, A is indifferent between y
and x.
(2) A’s preferences among options are transitive.
-5(3) A seeks to maximize his or her utility where the utility of an option x is
greater than the utility of an option y for A and if and only if A prefers x to y.
The utilities of options are equal just in case the agent is indifferent between
them.
(4) If option x is acquiring commodity bundle x’ and option y is acquiring
commodity bundle y’ and y’ contains at least as much of the commodity as x’
and more of at least one commodity, then all agents prefer y to x.
(5) The marginal utility of a commodity c to an agent A is a decreasing function
of the quantity c that A has.
(6) When we increase any input into production, other things being equal, output
increase, but, after a certain point, at a decreasing rate.
(7) Increasing all the inputs into production in the same proportion increases
output by that proportion. The production set is weakly convex and additive.
(8) Entrepreneurs or firms attempt to maximize their profits.
Hausman says that many people would say that (1) - (3) assert that people are
rational and (1) - (4) assert that people are economically rational. He admits that this
equilibrium theory is rough and incomplete, and that “(7) and (8) are the most likely to be
dropped or replaced by contrary generalizations, while reveal preference theory is
supposed to supplant (1) – (5) … These characterizations of prevailing circumstances are
not regarded by economists as discovered by or as asserted by their theory. They provide
a sketch of the circumstances to which the theory is applied and are therefore crucial to
the derivation of important theorems, but they are not themselves assertions of the theory.
The term ‘assumption,’ of which economists are perhaps overly fond, fits these
stipulations well” (p. 346).
Replacing any one of (1) – (8) with a non-equivalent
generalization would count itself as a theory change. Most economists would concede
that (1) – (8) are not true (p. 356), and that these statements don’t exhaust the ‘laws’ of
microeconomics, but “the above eight lawlike statements express the fundamentals of
neo-classical economics” (p. 346).
-6From the perspective of the view of the prescriptive epistemic status of ‘axioms’
and certain definitions in formal systems, it is accurate to conclude that the lawlike
assumptions found in an economic equilibrium theory, are stipulated or prescribed as a
set, as the means for making microeconomic policy decisions. Hausman says that the
theory “can be extended to deal with questions of economic growth and change. It is the
theory which the economist relies on in empirical research and in many welfare
recommendations” (pp. 344-345). Not only are the axioms and certain definitions of
formal deductive systems prescriptive, the sets of axioms, assumptions, stipulations, and
definitions in various social science theories (e.g. rationality, consumer choice) are the
prescribed means for the measurement of certain economic and social phenomena. 3
III. 'Adequate Evidence' Assertions are Prescriptions
A third kind of prescription are assertions of 'justification,' 'sufficiency,' or
'adequacy' of evidence in a claim to have knowledge. Condition 4a in the PE definition
specifies what an 'adequate evidence' sentence is. Personal justifications end on a persondependent mode in the activity of justifying a belief. When S appeals to the background
premises for believing a particular premise as part of a longer argument, a regress of
reasons is terminated by persons who accept a given premise to be true without further
questions or doubt. There are no objective criteria for determining when a regress should
end. There are no objective weighing mechanisms to describe when given evidence is
'sufficient' for S to be 'justified' in believing p. Assertions that 'S has presented sufficient
premises for believing p' prescribes that there is no further need for inquiry about the
truth of p. In these situations, it is believed that there is a negligible possibility that there
is unconsidered evidence that might lead to genuine doubt that p.
3
Further evidence supporting the belief that measurement systems are constructed by persons includes the
fact of alternative systems of measurement for identical domains. Three examples come to mind. The
assertion 'This stick shall be the standard length of one meter' when uttered centuries ago (referring to a
particular stick) to define the standard length of a 'meter' was prescribed as a stipulative definition. That
there evolved a metric system using centimeters, decimeters, meters, and kilometers, and (at the same time)
a weights and measurement system using inches, feet, yards, and miles as standard units indicates that these
systems have prescriptive axioms and definitions. Similarly, that there is a Fahrenheit and Celsius scale for
measuring temperature is evidence that the axioms that underlie these two systems are neither true nor
false, nor contradict each other. With regard to contemporary physical theories, assertions within quantum
mechanics can be equally-well stated with an alternative ontology of either 'particles' or 'waves.'
-7IV. Normative Ethical Assertions Are Prescriptions
In chapter five, I introduced the following metaethical theory:
Prescriptivism: Ethical assertions and substantive value affirmations are
prescriptions. The 'correctness' of any value affirmation or ethical
assertion is dependent upon what persons accept, tolerate, or agree-to, and
does not refer to an objective moral reality.
Prescriptivism maintains that ethical assertions may be approved (or disapproved) by
humans, but that they are neither true nor false. Any argument with an ethical 'ought'
conclusion is always derived from a set of premises that includes at least one prescriptive
(ought) assertion. A moral argument includes prescribed value(s) or ethical principle(s)
in its premises, in addition to facts, to support a moral conclusion. Prescriptivists speak
of normative conduct as being 'right' or 'wrong' but without pretense that there is
something morally objective that 'lies beyond' the values that we endorse.
The Frege-Geach Problem
Much attention has been paid to the so-called Frege-Geach problem that has been
widely-discussed over the past several decades (e.g. see Mark Schroeder 2010). Peter
Geach (1965) alleges that there is a theoretical problem in understanding how
prescriptive premises (having no truth conditions) can function to produce valid moral
arguments. Geach's primary problem with non-cognitive theories of moral assertion is
that they do not specify any truth-conditions for a moral assertion. This allegedly presents
a problem of equivocation when evaluating deductive arguments. Geach, in effect, asks
how can a prescription carry consistent semantic contents across both asserted and nonasserted contexts and where contexts differ. We will review two problems of possible
equivocation in ethical deductive arguments.
Problem #1: The Indeterminate Value of the Antecedent in Premise 2
This example is from Geach (1965, p. 463):
1) Tormenting the cat is bad.
2) If tormenting the cat is bad, getting your little brother to do it is bad.
Thus: 3) Getting your little brother to torment the cat is bad.
-8Adjusting Geach's problem to prescriptivism, it is alleged that if one assumes that the first
premise is prescriptive, it isn't clear in the second premise whether all persons would
accept or reject the antecedent prescription. If the antecedent in the second premise has
undetermined acceptance or non-acceptance because 'tormenting the cat is bad' in the first
premise has either a positive or negative prescriptive acceptance, then we are
representing 'tormenting the cat is bad' in the antecedent with potentially two different
acceptance-values, and the above argument is invalid on the pain of equivocation.
Our response to this problem is that we should just charitably assume that the
second premise has the form 'if prescription, then prescription,' and that the antecedent of
the conditional (in the second premise) and the first premise have the same-acceptance
value, and that the moral conclusion is prescriptive:
1) (It is prescribed that) Tormenting the cat is bad.
2) If (it is prescribed that) tormenting the cat is bad, then (it is prescribed that)
getting your little brother to do it is bad.
Thus: 3) (It is prescribed that) Getting your little brother to torment the cat is bad.
In this argument, we would reasonably assume that anyone who asserts the first premise
as an assumed-prescription would also assert the antecedent of the second premise as the
same prescription as a matter of consistency. As long as the prescriptivist who advances
this argument maintains that there is no literal truth value to the first premise, the entire
second premise, and the conclusion, then there seems to be no equivocation in meaning.
This moral argument seems just as unequivocal as the following modus ponens argument
that exclusively employs descriptions:
1) (It happens to be the case) it is raining.
2) If (it happens to be the case) it is raining, then (it is the case) that the streets
will be wet.
Thus: 3) (It is the case) the streets are wet.
In this non-moral argument, it is contingent whether the first premise (a
description) is truly the case (i.e. that it is raining at a time and place). Similarly, it is a
contingent in the first premise of the moral argument (about cats) as to whether it is
prescribed (by anyone) that tormenting the cat is wrong (and whether a cat's life is
-9valued). With this ‘raining-streets example’ it is illustrated that when the antecedent of
the second premise is assumed to be (contingently) true, the sentence has a truth value.
The assumed-truth of the contingency of the antecedent, in premise two, indicates that the
premise is indeed appropriately classified as being a (true or false) descriptive assertion.
In fact, premise two is a (true) proposition based upon a belief about physical causation.
JC Beall and Greg Restall (2006) state Frege and Geach’s problem about the
‘indeterminate value of the antecedent’ of premise two with the following example:
1) Zack will sleep well tonight.
2) If Zack sleeps well tonight, he will be happy tomorrow.
Thus: 3) Zack will be happy tomorrow.
According to Beall and Restall, although this argument is an apparently valid inference;
the alleged problem is that in the second premise, the antecedent “not asserted at all” (p.
10). What is the relationship between this first premise which is a true or false assertion,
and the second premise which isn’t an assertion? Using a formula similar to above, and
adding a context it is easy to interpret the second premise as a descriptive assertion:
1) Zack will sleep well tonight (e.g. known as a causal descriptive assertion,
based upon Zack’s recent act of taking sufficient sleeping pills).
2) If (it is known) Zac sleeps well tonight, he will be happy tomorrow (i.e. the
entire conditional sentence is a fallible causal-correlative descriptive assertion,
based upon the typical psychological dispositions of well-rested persons).
Thus: 3) Zack will be happy tomorrow.
With an ordinary contextual interpretation, the antecedent of premise two is asserted as
known, and is either true or false. It is false that this antecedent is ‘not asserted at all’ in
an ordinary context. Moreover, the consequent of premise two is asserted as a descriptive
predictive assertion. The well-formed second premise in this argument is a descriptive
assertion (purportedly true, but insufficient for next day happiness) and doesn’t affect the
obvious validity of this deductive argument. If we take the components of deductive
arguments to be sentences/propositions asserted in contexts-of-use, there is no FregeGeach problem.
-10Problem #2: The Indeterminate Value of the Consequent in Premise 2
Another problem is illustrated with the following similar moral argument:
1) Feeding the wolves is bad.
2) If feeding the wolves is bad, getting your little brother to do it is bad.
Thus: 3) Getting your little brother to feed the wolves is bad.
In a context, it could be the case that feeding wolves (e.g. in a residential neighborhood)
is bad because of ills associated with the congregation of wild wolves among humans. In
this situation, the alleged entailment to a moral conclusion might be thought to be invalid
because even if premise 1 and the antecedent in premise 2 are consistently adopted (as
assumed-true or prescribed), this leaves open as indeterminate the consequent about
whether you should get your little brother to do something that is bad.
Since the
consequent is a separate undetermined value judgment, this leaves the conditional in
premise 2 with an undetermined truth/adoption value, and the argument isn't valid.
The solution to this problem is to just simply make explicit (i.e. add) an already
implicit suppressed third premise:
1) (It is prescribed that) Feeding the wolves is bad.
2) If (it is prescribed that) feeding the wolves is bad, then (it is prescribed that)
getting your little brother to do it is bad.
3) (It is prescribed that) Getting your little brother to do bad things is bad.
4) Thus (It is prescribed) Getting your little brother to feed the wolves is bad. 4
To repeat from above: With a moral argument, the best we can do is to present a valid
argument, where it is assumed (as a fiction) that the value and ethical premises have a
truth value, and that the validity of the argument is determined by the standard rules of
deductive logic. The assumption that ethical values and principles are either 'true or false'
is false, but there is no harm in assessing the validity of moral arguments, as long as it is
understood that the value premises don't literally have a truth value.
When using
prescriptions in arguments about the proper treatment of cats and wolves there is no
4
As an aside, in other contexts, such as at a public zoo, where food is provided by the zoo, it could be the
case that it is permissible (or good) to feed the wolves. The sentence 'Either feeding the wolves is bad, or it
is not' (in premise 1) cannot be evaluated out of context.
-11equivocation in a speaker's meaning and intent. Similarly, in the argument about drunk
driving there are no equivocations.
Metaethical Prescriptivism Compared to Expressivism
Expressivism is a family of non-cognitive metaethical theories that characterizes
ethical assertions as expressing non-belief-like states of a person's attitude (or state of
mind). When making a moral assertion, one is not just voicing one's own attitude, but
one is also seeking to influence the attitude of others. An attitude is described as a
practical stance or policy of action. The utterance of a moral assertion has the function of
coordinating actions if adopted by others. Moral words are used to express the emotions,
feelings, or attitudes of a speaker. A.J. Ayer (1946), Charles Stevenson (1944), Simon
Blackburn (1993, 1998), and Allan Gibbard (1990, 2003) are all of the non-cognitive
expressivist tradition.
The prescriptivist response to expressivism is to modify (and reject) the
expressivist thesis that moral assertions function to express attitudes (i.e. non-belief
mental states). Instead the prescriptivist understands moral assertions to function to
express a stipulation (or rule) upon a practice, where the correctness (or incorrectness) of
the assertion is dependent upon its acceptance (or non-acceptance) by particular persons.
Moreover, the prescriptivist doesn't posit that moral words (e.g. wrong, good) have a
different kind of meaning than non-moral words. Rather the attention is to a speaker's
intent and the contextual meaning of complete sentences.
In claiming that moral
assertions should be understood as being prescriptive, the prescriptivist identifies the
'non-belief attitudes' of the expressivist as one's personal values. 5
5
Values are physically manifested in sentient creatures the same as beliefs and concepts are manifested in
sentient creatures. Value utterances about morality are characteristically expressed as well-formed
grammatical sentences the same as statements of belief. But value is internal and owes its existence to the
interests, desires, and attitudes of humans (and other sentient creatures). Values can be changed or adjusted
on the basis of new facts, or with sensitivity to value conflicts or differences in value. The willingness to
consistently follow one's own values (i.e. having moral motivation) is variable internally with respect to
particular issues (and situations) and among persons.
-12The Limits of the Knowable
Given that there are several kinds of prescriptive assertions, a conclusion
pertaining to the limits of knowledge, is that prescriptive assertions are not true nor
false, nor knowable. In contrast, a descriptive assertion is actually correct (i.e. true) and
potentially knowable only if its content corresponds to (or reflects) a state of affairs. The
definition of knowledge and descriptive-prescriptive thesis do not endorse ‘empiricism’
(roughly, the theory that all knowledge is obtained by the five senses).
Instead, a
substantial amount of knowledge comes from understanding the intentions motivating
natural and artificial languages, and not directly from sense experience.
I. Empirical Assertions Are Descriptions
Statements about material affairs are the most common form of description.
Sentences about empirical matters (e.g. 'I see a table,' 'Chicago is sixty miles south of
Milwaukee,' or 'My knee hurts') are mundane and there is no question that they are
descriptive in meaning and have an objective truth value in a context. Theories in the
physical sciences involve some stipulations and prescriptive assertions, but the content of
scientific empirical theories are mostly descriptive.
II. Analytic Sentences, Deductive Entailments, Tautologies, and Applied Deductive
Propositions Are Descriptions
True propositions of mathematics and symbolic logic are derived relative to the
stipulated foundations (axioms, definitions, inference rules, grammar, and vocabulary) of
a deductive system. The derived propositions that 'three is prime,' '7+5=12,' 'the sum of
the interior angles of a triangle is 180 degrees,' and 'Abraham Lincoln is Abraham
Lincoln' are true-in-a-deductive-language necessities.
Thus, when applying standard
measurement systems, the truth of these statements expresses a correspondence (or a
reflection) of some state of affairs, where the correctness (or incorrectness) of the
statement is independent of its acceptance (or non-acceptance) by particular persons.
Mathematical propositions when applied to practical matters can be objectively true: If
Sam has 1 marble and Suzie has 8 marbles, then they have 9 marbles in total.
-13III. & IV. Reportive and Theoretic Definitions Are Descriptions
These two categories of assertion have a linguistic entity (i.e. a definiendum) as
their subject. Reportive definitions (truly or falsely) describe linguistic practice. A
definiendum is claimed to have a standard definiens in a given language or language
community. Theoretic definitions purport to describe the nature of an entity or
phenomena represented by a definiendum, with the intent to make the definiendum-todefiniens relationship a true one.
V. Aesthetic Assertions Are Descriptions
It may seem a little surprising, but in chapter eight, it was argued that another
kind of descriptive assertion is that of aesthetic judgment. When a person states that 'U2
is the best contemporary band' or that 'the painting is beautiful' or that 'The Horse
Whisperer was a good movie,' that person attributes a (favorable) relation between her
preferences (or values) as an existing mental state to entities which satisfy those
preferences. A person is describing what she likes. In most cases, we grant that persons
can know what aesthetic experiences (truly) please them. Because there is some intersubjective commonality of aesthetic value, prescriptive recommendations often follow
from aesthetic experience, e.g. whether a movie is worth seeing, or whether a kind of ice
cream tastes good, and so on.
VI. Social Science Theories Are Descriptions
The social sciences include Anthropology, Economics, Education, Geography,
History, Linguistics, Political Science, Psychology and Sociology, among others. Social
sciences draw upon empirical methods and attempt to be objective, and to emulate the
practices of the physical sciences. In the social sciences, scholars seek to describe
expectations of how persons will behave on the basis of the beliefs and desires attributed
to them.
Social science theories explain beliefs and behavior by rendering them
intelligible. Models and theories attempt to simulate a world that explains human
intentions. In chapter eleven it was suggested that analytic philosophy should likewise be
-14conceived of as a 'social science' on a group resemblance basis. Philosophers engaging
in conceptual analysis should investigate linguistic evidence, intuitions, and speaker
intentions when presenting an abductive argument supporting a philosophical thesis.
A 'Prescription' Exceeds the Explanatory Power of a 'Performative'
The concept of a 'prescription' supplants J.L. Austin's (1975) notion of a
'performative' utterance. Austin introduced the performative as a kind of assertion that
isn't true or false, but is used to do something (to promise, to warn, to bequeath, to bet,
etc.). Austin's performatives that are 'used to do something,' has had little theoretical
impact. Austin's examples can be subsumed in terms of descriptions and prescriptions:
1) An 'interrogative' (e.g. 'Do you know where a gas station is?') is interpreted as
the conjunction of a description and prescription: 'I do not know x' (description)
and 'please tell me x' (prescription).
2) The concept of a 'promise' is to sincerely describe one's intention to do
something, and to prescribe to oneself to perform appropriate follow-through
actions.
3) With a 'warning' (e.g. 'Watch out!') a prescription is asserted, often
accompanied by a description ('You'll get hit') about probable consequences of
not heeding a warning.
4) In 'bequeathing' to assert 'I give and bequeath my wristwatch to my brother,
after I die' is to describe one's wishes, and prescribe to executors to abide by
one's will.
5) The 'solicitation of a bet' (e.g. 'I'll bet you $25 that the Green Bay Packers will
win') describes a bettor's willingness to bet money on his belief (prediction) about
the outcome of a sporting event, and prescribes to the listener to accept the wager.
6) A 'request' (e.g. 'Would you please close the door?') is a prescription that a
person should aid the speaker, and implicitly describes that the speaker desires (or
has value) in having the door closed.
7) Whether a sentence is being used to describe or prescribe (or both) can be
relative to social context. For example, a cook speaking to a lunch patron asserts
'Your sandwich is ready' describes the fact of the sandwich's completion, and
prescribes patron pick-up. The assertion 'In order to turn off the lights you must
flip the switch' can be ambiguous without context. The speaker may be informing
the hearer about how to turn off the lights (i.e. describing) in a room or the
speaker may be requesting the hearer to turn off the lights (i.e. prescribing).
-15A Comparison with the Declarative-Interrogative-Imperative Distinction
With respect to the importance of the descriptive-prescriptive distinction about
speaker meaning, it can be contrasted with what Sadock and Zwicky (1985) observe are
the three basic sentence types with similar functions in most natural languages:
These are the declarative, interrogative, and imperative. At first approximation,
these three types can be described as follows: The declarative is used for making
announcements, stating conclusions, making claims, relating stories and so on.
The interrogative elicits a verbal response from the addressee. It is used
principally to gain information. The imperative indicates the speaker's desire to
influence future events. It is of service in making requests, giving orders, making
suggestions and the like (p. 160).
One obvious problem with this functional classification of sentence meaning is that
imperatives are often asserted as declaratives (e.g. 'It is wrong to do x'). In comparison,
the descriptive-prescriptive theory of speaker meaning allows for more philosophically
insightful explanations than does the declarative-interrogative-imperative description of
speaker meaning.
Conclusion
In the overview, this book focuses upon three areas of concern: (1) ‘concepts’ (as
being physically instantiated within our brains), (2) the nature of natural and artificial
languages (including a ‘theory of definition’ and the descriptive-prescriptive speaker
meaning distinction), and (3) an explanation of the relationships between human beliefs,
values, and intentions in the (successful) attainment of ‘knowledge’ while assuming a
physicalist ontology (and a social scientific methodology). Our questions have been:
What is knowledge? What is knowable?
With the recent introduction of naturalized epistemology, reliabilism, and virtue
epistemology, there has been an abandonment of any attempts at necessary and sufficient
conditions analyses of knowledge. In response, it has been shown here that knowledge
can indeed be analyzed as a theoretic natural kind entity. A theory of knowledge should
be about beliefs, ordinary language assertions, and about what we can know. The
predominately externalist definition concisely states what knowledge is, and provides a
theory for understanding what kinds of assertions can be known (and not known). It is
-16concluded here that (1) stipulative definitions, (2) the axioms and inference rules of
formal deductive and metaphysical systems, (3) 'sufficient evidence' assertions, and (4)
normative ethical assertions are prescriptions and cannot replace p in 'S knows p.'
In
contrast, (1) empirical statements, (2) true-in-a-language sentences (i.e. analytic
sentences, valid entailments, and tautologies), (3) applied mathematical/deductive
assertions, (4) reportive definitions, (5) theoretic definitions, (6) aesthetic assertions and
(7) social science theories are descriptions, and are potentially true and knowable.