Today's class will be split roughly in two equal parts.  In the first half, we cover Hollis' Philosophy of Social Science, working through his arguments about:
    a)  understanding vs. explanation
    b)  holism vs. individualism
    c)  "epistemological warrants" or how we can justify knowing what we think we know, where we cover points about rationalist theory, empiricist theory, and a blend of the two based on ideas from Popper, W.V. Quite, and Kuhn.

The second half of the class turns to Sociological classics on questions of theory and scientific method.  We start w. texts on method per se (Durkheim and Merton), and end with questions about the moral / political scope of social theory (Weber's "Science as a Vocation" and "Objectivity and Social Science").
 

I.  The Philosophy of Social Science

I.1 Background.

1. School book image of science is of unprejudiced Reason explaining nature.
2. 17th century observers showed that the world didn't work like accepted texts showed that it did.  This lead us to a problem of method: how do we identify what is happening in the world?
3. Call the method they developed, "Reason". Reason does three things:
• Brings to light new matters of fact
• extends facts we know to things we don't know
• the inner workings of nature are revealed through Laws
4. The method was turned on society, in both it's positive and normative guises.

To start, there is a basic distinction between explaining the world TOP DOWN vs BOTTOM UP.

I.2.  Structure and Action.
    The example that Hollis uses here is Marx, specifically his Preface to a Contribution to the Critique of Political Economy.   The world is driven by Forces, saying that

"It is not the consciousness of men that determines their being, but, on the contrary, their social being that determines their consciousness."

1. an Ontology - a set of assumptions about 'what is out there' to be explained.
2. a methodology -A way of organizing evidence to find answers, that can justify movements from statements of fact to theoretical statements.
3. an epistemology - A theory of Knowledge:
• how can we know anything at all?  What counts as justifiable knowledge and how do we judge a true account from a false account?
• Does our knowledge of ourselves (social knowledge) differ in a fundamental form from our knowledge of the rest of the natural world?

• Ontology:  Pressures, Systems, Forces that underlie social life and individual action.  A world of unknowable forces
• Methodology:  A naturalistic (rational), scientific method
• Epistemology: A rationalist epistemology that uses theory / logic to identify necessary and unseen facts.

In contrast, is a "Bottom-Up" approach, which Hollis identifies through the Mill example.  In On Liberty, Mill argues that in an open society, people are critical thinkers who drive social progress. (see p.10).  Human being have no properties except as people, and social science must be grounded in the laws of "individual man"

"The laws of the phenomena of society are, and can be, nothing but the laws and actions and passions of human beings united together in their social state."

For Mill

• Ontology: Only individuals, their possessions or actions
• Methodology: Very similar to Marx: natural science used to find covering laws.
• Epistemology: Empiricist, you can only know what you see/observe.

Hollis now introduces a useful heuristic for organizing social theory:
 

	Explanation	Understanding
Holism	Systems  (Structure)	Games  "Culture"
Individualism	Agents  (Action)	Actors  (Role players)

The Holism - Individualism dimension distinguishes Marx from Mill in the example.  Holists (what might, in other texts, be called "structural", though he avoids this term for good reason) seek to understand the social world at a level onto itself.  If we are concerned with individual action, it is as an outcome resulting from system behavior (people's consciousness are the result of historical forces, for example).  Individualists argue that anything that is social must, in some way or another, be reduced to properties of individuals (it is, I think, an open question where systems of interaction fall, as we will see).

• SYSTEMS "Systems" refers to theories that see the world (society) as a phenomena to be explained as a function of itself, not as a reference to lower level (i.e. individual) behaviors.  Durkheim's theory of social transformation fits this cell.
• GAMES Refers also to a macro, emergent, holistic approach, but from the standpoint of UNDERSTANDING instead of as as natural, necessary point of explanation.  Cell 2 is referred to (by Hollis) as 'games' because what is often studied in this cell is how rules for behavior, normative structures that guide how people interpret and move through the world interact with each other and transform/evolve. This is what we might (somewhat sloppily) call Culture -- the system of rules and meanings that comprise social life.  Note that this is not an un-controversial designation.  Durkheim discusses the stuff of culture (oaths, marriages, etc.) in a systems way.
• AGENTS.  Refers to a micro, bottom up approach to understanding the social world.  Here we take as our starting point the actions and behaviors of individuals and attempt to explain why certain structural characteristics exist.  We attempt to account for the whole as a (knowable) function of the parts.
• ACTORS again refers to the micro level of analysis, taking as primary the actions of individuals.  But says that the whole, the superstructure, is only an outcome of the many individual interactions and meaning episodes.  Thus, the play isn't a script written beforehand, but an impromptu theatrical adventure guided only by past actions and the constraints of other actors.  Another analogy for this cell is a Jazz band - - good jazz follows from the ability of players to take a sketchy line and create a new musical 'whole' through the back and forth of good players.

 Understanding vs Explanation.
          Many social thinkers dispute the enlightenment ideal that society is simply another facet of nature that can be 'explained' -- like astronomy or even biology.  Instead, society needs to be understood from within, as a system of meanings people don't just behave, or follow blindly a set of necessary constraints.  Instead actions are multi-valent, carriers of meaning: and any one 'act' is interpretable in multitudes of ways.  Thus instead of seeking the causes of behaviors, they look for the meaning of actions.
          Many thinkers cross these bounds with various parts of their work: Marx in the early manuscript is looking at the meaning of labor, and alienation falls out of an interpretive notion of work.  His analysis of capitalism, however, is much more an explanation, positing rules and laws.

Question:  What does "meaning", mean?  What, exactly, is involved in an "interpretive understanding" of social life?

Hollis is content to leave it a little ambiguous.  The key is that you attempt to make sense of social life using the same, arguably uniquely, social factors that people use themselves.  Concepts such as "value", "purpose", "development", "Ideal".

A systemic view that rests on understanding looks at social life through 'games', that set the rules for actors.  These are things that both guide action, and give action meaning.

Question:  Is the distinction between Understanding and Explanation the same as the distinction between the Humanities and the Sciences?

Determinism
    Throughout this book, Hollis has a minor sub-theme that we will want to keep in mind, especially if we are committed to a naturalistic, scientific source of explanation.  If there is a necessary causal effect active in social life, in what sense are people free to act?  Hollis sets this up as being a potential problem for both rationalist and empiricist traditions.
    For Marx:
        Economic Conditions --> Consciousness --> Action : implies no will.
    For Mill:
        Fixed human nature --> Fixed outcomes :  Implies no free will.

PERHAPS, any causal argument that rests on necessity, will have this outcome?  Determinism says that there is a complete causal order in nature, every event at time t is an outcome from a cause at time t-1.

ON THE OTHER HAND, if there is NO cause-effect relation, would human action be any better off?  Could one effectively act in the world, if you could not, in some strict sense, cause something else to happen?  This is the basis of Mills' argument that freedom presupposes a causal order.

We will not worry too much about this, other than to note that there is a continuing tension between structural explanations of social life and Action explanations, which trip around this dilemma. Any strong theory of structure runs the risk of making action impossible, while any strong theory of action makes regular structure ephemeral.  The trick is to find a balance point.

Hollis starts his examination of the underpinnings of social science by working through 2 types of explanation:  the "rationalist" way ("bacon's first way") and the "empiricist" way.   In both cases, the method used for science might be thought of as a way to answer the question, "Why do you need a theory?"  The answer rests on some deep assumptions about what we can know.

Chapter 2.  A Rationalist way of discovering the truth?
The rationalist way, quoting Bacon, is:

"There are and can be only two ways of searching and discovering truth.  The one flies from the senses and particulars to the most general axioms, and from these principles, the truth of which it takes for settled and immovable, proceeds to judgment and the discover of middle axioms."

What does this mean?  First, it implies that there is a theory of knowledge that does not depend solely on observation.  For example, newton saw apples fall, but could only reason to gravity.  There are things that we can't see, like the springs in a watch, that 'really' cause the events of the world.  The purpose of theory is to make it possible for us to understand these things we cannot observe directly.  Bacon's first way is a search for universal laws that hold necessarily.

Appearance vs. Reality
In this model, appearance is the result of nature acting on the mind of an observer.  This makes it possible for appearance to be other than reality, which leads to a duality between reality and appearance.

This raises the key question. How can we know of something that, by definition, is unobservable?

The rationalist answer is that we know through the faculties of our mind.  We know the truths of Geometry, for example, through reflection.  This allows us to know, deductively, truth without experience.

Hollis is quick to point out, however, that while it may be possible to derive theorems from axioms, it is quite another to prove the theorem true.  To prove a theorem follows is not to prove that the theorem is true, unless one already knows that the premises [note the plural, many are implied and often unnoticed] of the proof are true.  Whenever there is more at stake than a coherent system (like math), coherence itself ceases to be a guarantee of truth. (It is often pointed out, for example, that Alice in Wonderland is perfectly logically consistent]

Science on this model works through middle axioms.  Some point of departure is needed, with the basic axioms taken as true.

SO, why does science need a theory?  Two answers:

1. Science is a search for cause, but observation only gives you correlation

To explain an event is to identify its cause, placing it in a series of events, each of which gives rise to the next.  Theory gives us necessity (the opposite is embodied in the phrase, "The senses know nothing of necessity")
 
2. Science is a search for universal laws, but observation only gives you particulars

Science seeks to find something that is true in all possible worlds.

Hollis argues that it is a mistake to equate logical necessity with causal (in-the-world) necessity.  BUT, if we reject the notion that logical necessity and causal necessity are equal, but want to remain realist, we will need a model for what 'cause' means.

Epistemologically, do we have a faculty of reflective reason, which lets us know what the senses cannot possibly tell us? [See Kant's Critique of Pure Reason here].

Chapter 3.  Positive Science: The Empiricist way
From Bacon, again:

    "The other [way] derives axioms from the senses and particulars, rising by gradual and unbroken assent, so that it arrives at the most general axioms last of all.  This is the true way, but as yet untried" (p.23)

Positivism
    [we should note, as Hollis hints, that "Positivist" is really a pejorative term in sociology, usually used to describe someone who is unthinkingly quantitative.  Very few people fit that model, so we are using the term in a much less loaded way, to refer to people who base their theory on empirical observation.]

Positive science, "goes with an empiricism about scientific knowledge, which rests on observations as the moment of truth when hypotheses are tested against facts of the world." (p.42)

Science consists of a body of beliefs some of which we know to e true and others of which we are rationally entitled to hold, given what we know.  The basic beliefs are warranted by perception, which is our only source of direct acquaintance with the world and hence our only warrant for basic statements about it. (p.43)

If x% of known As have the property B, then the probability that in the same conditions the next A will have the property B is x%.

The way out of this conundrum, for pure empiricists, is to claim that we use observations to test the theory we develop.  That is, we use something other than observation to develop theory (like induction), but always test the induction against the 'facts'.

Hume and Causation
At this point, Hollis introduces one of the trickiest problems out there: what, exactly, is "cause" and how can we know it, using Hume's classic treatment.

We may WANT more than perception, but logically, can we have anything else?  According to Hume, we cannot know cause observationally.  Instead, cause is simply,

"An object precedent and contiguous to another and where all objects resembling the former are placed in like relations of precedency and contiguity to those objects that resemble the latter."

What distinguishes cause, then, from coincidence?  Experientially, nothing.  General probabilistic statements are the only thing you can have, based on observation.

Positive Economics
Hollis uses Friedman's work on economics as an example. What matters is not the "reality" of a theories assumptions, but simply how well it predicts behavior.  Market assumptions are not realistic, but they are effective at predicting behavior.

[But I can't resists:  "The task is to provide a system of generalizations that can be used to make correct predictions about the consequences of an action"  Note the language, the action generates consequences, which, must, have a stronger sense of 'cause' underlying it.  See Judea Pearl's book "Causality" for a take on this]

The Analytic-Synthetic Distinction
Hume divided the world between "matters of fact" and "relations of ideas"  The world just IS.  Any natural necessity is something we bring to the world, is is beyond our ken to observe it.  Any statements which are necessarily true are not about the world, but about relations among ideas.  They depend solely on language, like math.

Analytic:  True as a function of logic and meaning of words. Example: 2+2 = 4, or "All bachelors are unmarried"

Synthetic: Truth or Falsity relies on matters of fact.

One must never confuse words w. things, according to MH, like thinking that it is a fact of the world that all bachelors are unmarried. This is simply "true by convention" and depends on the decision made by people (about the meaning of the words).  A more clear example: find a circle w. diameter other than 2pr, and you don't have a circle.

PUre theory, for friedman, is only a set of statements, a filing system.  Synthetic statements are reserved for hypotheses.

Does this mean that the Analytic statements are uninteresting? No.  Math, for example, is full of discovery and interest, but it "tells us nothing about the world".
[do you buy this? What would an engineer say about this kind of claim? Is that a fair, logical, retort?]

MH sums up w. 4 basic points on this chapter so far.  We have:

1. Basic empiricism is the key to science, which rests on (a) perception and (b) induction.
2. An analysis of "cause" that purges it of natural necessity (through the synthetic / analytic distinction)
3. An example through Positive Economics from M. Friedman
4. An epistemologically guided distinction between language and fact.

1. Rational Actors
2. Complete and constant preferences
3. All relevant information
4. Perfect ability to calculate

Are these analytic statements or synthetic statements?  MH says we can't really say.  ON the one hand, any preference can be squared w. reality ex post. As such, it is true by convention and an analytic statement. But many economists treat it as a statement about people (and many sociologists don't like it!).

Discovery and Validation
How damaging is it to give theory an active role?  THe objection was that no science could confine itself to generalization from observation.  Even science need richer assumptions.  With out this break, we would not have science.  A way around this hard empiricist route is to segment observation and induction in two steps.  On the one hand, we are free to "Dabble" in whatever ideas we may have about how the world works (or "as if" statements about how the world works), but we must always return to observation for validation.

See the two figures on methods for science as examples.

Conclusion
The type of theory discussed in this chapter is often referred to as the "hypothetico-Deductive" model (see Blau's formal work for an example in sociology).

The Ontology is of particulars.  Things we observe.  The Methodology is induction, using observation to identify regularity in behavior.  The Epistemology is a basic and simple version of empiricism.

THis leaves us with a lot to yet understand.  There's not much, really, to say for this brash empiricism, if you think that there is more to explanation than simple prediction.

Chapter 4 (not assigned, but we want to cover some of these points).
There are 3 main arguments that we need from this chapter to make sense of a basic method/philosophy of explanation (as opposed to understanding) in social science.  The goal is to find something between the two ways thus far discussed.
 

1. Logic of Falsification, from Popper
2. Webs of belief (from W.V. Quine)
3. Paradigms (from Kuhn).

Foundations.
Before we start, MH walks us through the basic nub of the problem, which rests on foundations of knowledge.  The basic problem of epistemology is figuring out what we can know.  Foundational approaches to the problem point out that if you were to make a list of all the things you know, and link them together, at some point the chain of inferences will (a) fall onto something you know for sure or (b) form a self-contained system, all of which are uncertain.  Having a foundation requires two things: (1) a set of certain facts and (2) a set of rules for linking those facts together.  We usually take both of these, at some level, to simply be self-evident.

Interpretation
Empiricism is vulnerable to the claim that truth is never prior to interpretation.  There are no "obvious, clear" facts.  Every thing we know, we know because of a set of concepts that shape our understanding.  These concepts are supplied by the mind (though society?).  Keep this in mind as you consider each of the 3 modern claims about knowledge.

Science as conjecture and refutation.
The logic of induction that we discussed above seems clear: you use a method (inductive reasoning) to extend what you observe. If your observations are consistent with this conjecture, then you keep extending.  What is the problem with this?

Popper asked, what makes a theory scientific?
    The traditional answer is that scientific theory is confirmed by empirical evidence..  The problem is that many theories can find supporting evidence, not because they are right, but because they are stated such that *any* evidence would support them.  The key to a scientific theory, is that it could be shown false.

A scientific method that is based on induction, saying that "The more As which are B, the more likely A=B", is not logical.  COnsider:

    (1) H --> O
    (2) O
    (3) therefor H

Is not valid.  COnsider the case where H = "It rains" and O = "My roof is wet". THen "If it rains, my roof is wet.  My roof is wet, thus it rained.  Is not true.  I may have hosed it down.

ON the other hand, if:

(1) H --> 0
(2) not-O
(3) therefore not-H,
IS a valid inference.  This is the difference between confirmation and falsification.

THis is the basic logic used among practicing scientists, social or physical, notwithstanding the critiques that are leveled at it.  You need to understand this logic to practice in this discipline.

As a critique, the decisive moment of falsification can't happen without an induction, which is a bit of a logical circle.  That is, if you falsify the theory once, why should we expect that it will be falsified in the next round?  The event "The theory was falsified" is a lot like "I see another white swan."  IF we are not justified in concluding that all swans are white from observation, then why are we justified in assuming the theory is false?

Furthermore, the falsification rests on observation, which may always be theoretically informed.

PART of the answer lies in the method used to falsify the theory.  We use methods, ideally, that are designed to be regular and exact, such that there is a high likelihood that what one experimenter does, another will replicate (or, in principle, can replicate).

While this is the modern method of science, it is not without critique. Two prominent critiques follow.

Science as a web of belief (W.V. Quine)
Quine argued that two dogmas are at the heart of logical positivism.  First, the distinction between analytic and synthetic statements.  Second, the uninterpreted character of observation.

What distinguishes an "analytic" statement from a "Synthetic" statement?  Quine argues that it is words, such as 'necessity' and and 'logical', which have meanings that are open to revision, based on discovery and observation (if they mean anything else, then they must depend on some notions that can't fit with empiricism, like a 'real' unknowable world).  If, instead, analytic statements are open to discussion and revision, this distinction is not as valid as Logical Positivists might hope. AND making logical 'truth' open to revision opens a very thorny set of issues.

SEcond, the meaning of terms come from how they are embedded in in webs of other terms.  So too are most logical connections, including theoretical statements. This web provides defenses for the non-conforming observations. We can choose which parts of the theory are rejected by the evidence.

By deciding what observations are, we may be deciding between rival theories before we ever start.  [think, for example, of behavior genetic models].

Paradigms and After
Why do we see the world the way we do?  THere are (at least) two answers.
1) that we are hard-wired to do so (see Quite and Kant)
2) that cultural ideas shape our understanding.
    With this notion, epistemology comes very close to falling under a sub-field of sociology...

Kuhn argued for a distinction between Normal and Revolutionary science.

Normal science rests on paradigms, sets of shared ideas that guide our understanding of observations.  There are two things that drive paradigms:

1) Intellectual: A set of guiding axioms that we take as self-evident
2) Institutional: Normal science is embedded in practices that favor the status quo, through social mechanisms.

MOST of science is done within the bounds of a paradigm, making small advances in problem solving.  From time to time, however, a set of observations no long conform to the theory (Be aware, here of thorny issues about what data mean), and you get a crisis, which changes the paradigm and what we think about facts.

THere are many details to this argument.  The most interesting, perhaps, is the extent to which a crisis occurs because of objective facts that don't conform, or because of institutional changes in our understanding of facts.

Does sociology have a paradigm?  Do sub fields?  What are the implications for how we do science?
 

Part II.  Sociologists on how to do sociology.

1.  Durkheim.  Rules of Sociological Method (Follow link to get notes)

2.  Merton. Middle range theory & What Social Research has to contribute to theory

    This is one of the most famous works on method in the social sciences. Here Merton is arguing against the idea that we should build systems of social thought.  INstead, we should focus our attention on more confined empirical questions, theories of particular subjects, not grand theories of society.  He starts w. a very natural-science definition of theory:

MOST of the work in practicing sociology centers around middle range theory, concerned with family, crime, social balance, integration, etc.  You will need to understand (and, in 99% of all cases) apply Middle Range Theory ideas in your own MA and Ph.D. work.

Sociological theory refers to logically interconnected sets of propositions from which empirical uniformities can be derived.

Theories of the middle range are broader than working hypotheses but more focused than complete systems. It involves abstraction, but they are close enough to observed data to be incorporated in propositions that permit empirical testing.

Note, as well, the scope that Merton gives theory.  Theory should guide empirical investigation.  It should "generate distinctive problems for sociological inquiry"

These types of theories have not been derived from other, overarching, system-wide theories, and they are likely consistent w. many such theoretical systems.

Total Systems of Sociological Theory
Why not search for total systems?  Three reasons:
 

1. The first misinterpretation assumes that systems of thought can be effectively developed before a great mass of basic observations has been accumulated.  We simply haven't looked at the data long enough, yet, to accumulate these kinds of basic observations.  [as a good example of this, think of Astronomy's "main Sequence" scatter plot.]

1. That all cultural products existing at the same moment have the same degree of maturity.  Just because Physics is mature, doesn't mean that Sociology should be as well.  They have a 1000 year head start.

1. Sociologists often misread the state of theory in other disciplines.  Physics is not as unified as we might think from the outside.

The Bearing of Empirical Research on Sociological Theory
IN this work, Merton contrasts the typical Logical Positivist view that empirical research simply tests theory, and points out the multiple ways that empirical research builds theory.  This occurs in 4 principle ways:

1. The Serendipity Pattern

Finding new data and results pushes for new theory to explain them.  Fruitful empirical work originates new hypotheses.
Such data needs to be:
• Unanticipated
• Anomalous and surprising, usually because it is inconsistent w. prevailing theory/facts.
• Must be strategic, it must permit of implications which bear upon generalized theory, which rests more on what we bring to the data than in the observation itself.
His example is of the opportunity for baby-sitting in a community.  The real explanation rested not on population distribution, but on social connection.
 
2. The Recasting of theory

New data exert pressure for the elaboration of a conceptual scheme.
It is not that the data are anomalous or unexpected or incompatible, it is meerly that they had not been considered pertinent before.
Example: Danger and Magic
 
3. The refocusing of theoretic interest

New methods of research exert pressure for new foci of theoretic interest.
Examples: Statistical sampling, network methods, structural equation models, etc.
[is this good?]
 
4. The Clarification of concepts

Empirical research exerts pressure to clarify concepts. We need to have clear concepts to measure, and findings often point out how unclear our concepts are, which leads to clarification.

3.  Weber: Objectivity and the Social Sciences / Science as a Vocation

Objectivity in the Social Sciences
This selection outlines three of Weber's major points about how to do social theory.

1) Ideal Types
    Weber introduces the concept of an ideal type to be used in analysis.

• "The ideal typical concept will help to develop out skill in imputation in research:  it is no "hypothesis" but it offers guidance to the construction of hypotheses.  It is not a description of reality but it aims to give unambiguous means of expression to such a description. ... In its conceptual purity, this mental construct cannot be found empirically anywhere in reality. It is a utopia." (p171)

 

 

An ideal type is "...a matter of constructing relationships which our imagination accepts as plausibly motivated and hens as 'objectively possible' and which appear as adequate from the nomological standpoint." (p.172)

Contrast this idea w. what Hollis said:  we use something we don't observe, an ideal type, to think through how the world works. This is very similar to the "Second Way" of empiricism, if it were not for the focus on empathetic understanding.

"It is possible, or rather, it must be accepted as certain that numerous, indeed a very great many, utopias of this sort can be worked out, of which none is like another, and none of which can be observed in empirical reality as an actually existing economic system, but each of which however claims that it is a representation of the idea of capitalistic culture." (p.172)

Given that we can have many different types, how do we distinguish among them?  What use are they for an empirical science?

He points out that, in the practice of doing history (science), "the historian as soon as he attempts to go beyond the bare establishment of concrete relationships and to determine the cultural significance of even the simplest individual event in order to characterize it, must use concepts which are precisely and unambiguously definable only in the form of ideal types." (p.172)  Ex: "individualism", "Imperialism", etc.

"Every concept which is not purely classificatory diverges from reality."  THat is, science needs ideal types because science is not only interested in describing a given particular, but in abstraction, universality, and necessity.

"But social science in our sense is concerned w. practical significance. This significance however can very often be brought unambiguously to mind only by relating the empirical data to an ideal limiting case." (p,173)

"Nothing is more dangerous than the confusion of theory and history stemming from naturalistic prejudices." In 3 senses:

1. Confusion arising in the belief that the 'true' content and the essence of historical reality is portrayed in ideal types
2. in the use of ideal types as a bed into which history is to be forced
3. in the hypostatization of ideal types as real forces and as 'true' reality, which operates behind the observed facts [contrast this statement to Hollis' treatment of Marx]
• He uses examples from economics (capital accumulation, competition, etc.)

"These evaluative ideas are for their part empirically discoverable and analyzable as elements of meaningful human conduct, but their validity cannot be deduced from empirical data as such.  The "objectivity" of the social sciences depends rather on the fact that the empirical data are always related to those evaluative ideas which alone make them worth knowing and the significance of the empirical data is derived from these evaluative ideas." ([.176)

3) Objectivity.
    We should seek to analyze the world objectively, setting aside our political and moral beliefs.  HE says,

    "the elementary duty of scientific self-control and the only way to avoid serious and foolish blunders requires a sharp, precise distinction between the logically comparative analysis of reality by ideal types in the logical sense and the value-judgement of reality on the basis of ideas." (p.175).

Continuing, he argues an empiricist line:  that "..the value of those truths which empirical knowledge alone is able to give us." (p.175) sets up the same model Hollis is discussing, that we have to present ideas and test them against data.

Science as a Vocation
I'm not going to summarize the article in detail.  The first half is a fun read comparing German to American university life, with some points that simply haven't changed much since he wrote it.  The key point, for us, in this article rests on what a professional social scientist should do, with respect to politics.

Scientists, as scientists, should avoid giving their political and personal points of view, and stick to the business of doing science.  But, what does this mean?

Cut to p.143.  "...what is the meaning of science as a vocation, now after all these former illusions, the 'way to true being', ..., have been dispelled?  Tolstoi has given the simplest answer.  Science is meaningless because it gives us no answer to our question, the only question that is important to us, 'what shall we do and how should we live'?  That science does not give us an answer is indisputable.  The only question that remains is the sense in which science gives us 'no' answer, and whether or not science might yet be of some use to the one who puts the question correctly."

Here he throws down the challenge.  How does he answer?

Start by saying what science presupposes:
    a) the rules of logic and method
    b) that what is yielded by scientific work is important in the sense that it is 'worth being known'.

It is this last that gives us trouble, for there is no scientific way of know what this means.  It must be interpreted, with reference to ultimate meaning.
 

Weber says
   "One can only only demand of the teacher that he have the intellectual integrity to see that it is one thing to state facts, to determine mathematical or logical relations or the internal structure of cultural values, while it is another thing to answer questions of the value of culture and its individual contents and the question of how one should act in the cultural community and in political associations." (p.146)

Why not?  Because criticisms is impossible in the classroom.  "In the lecture room we stand opposite our audience, and it has to remain silent."

Does this mean that, in a discussion-based class, it is OK to express opinions?

"The primary task of a useful teacher is to teach his students to recognize inconvenient facts -- facts that are inconvenient for their party opinions." (p.147)

There are deeper, scientific, reasons for not mixing politics and science.

Scientific pleading is meaningless in principle because the various value spheres of the world stand in irreconcilable conflict with each other. ... "I do not know how one might wish to decide scientifically the value of French and German culture."

What, then, does science contribute to practical life?

1. Science contributes to the technology of controlling life by calculating external objects as well as man's activities.
2. Science contributes methods of thinking and tools of thought.
3. Science helps us gain clarity: the teacher can link your ends to the means, though analysis. The teacher confronts you with the choice, not the solution.