A Short Critique of Long's Theory of Constitutive Means

Economics Political Theory
21

Leonidia:

Thank you for your reply and acknowledgement of some of the problems I was trying to point out.

Before I address your counter-example, I want to be clear that I do not necessarily believe that Professor Long will want to make the changes or adjustments you propose or recommend. If we assume Professor Long has been thinking and writing about ethics theory for fifteen or twenty years or more, we have to assume that he has a fair grasp of the nuances of his own theory, and a certain capacity to recognize the implications of various adjustments to his theory and/or its explication. He may see implications of your proposed changes that you do not see. It’s easy for you to recommend changes to the presentation of Professor Long’s theory. You have less of a personal stake if the proposed changes should lead to further problems, and maybe even worse problems. At least I assume you have less of a personal stake.

Below are some of the problems I believe would still exist even if Professor Long were to adopt your proposed changes. Let’s call your proposed changes Proposed Changes #1. If the problems I point out below have merit and/or are valid, then it is possible you or someone else will want to advance Proposed Changes #2. But if this occurs, I say we are now essentially revising Long’s theory and/or its explication from what he himself would present. Then we’re arguing about the theory of various revisors of Long, and no longer about Long’s theory as he would conceive it or present it.

My proposed solution is that I outline what I consider to be some problems that still exist even with your Proposed Changes #1. If you think they have no merit and/or are not valid, then this means all Long has to do is revise his presentation as you recommend. If my criticism(s) has merit and/or is valid, this means we may need Proposed Changes #2. Then, we are heading away from Long’s theory as he would present it. In either case, we needn’t debate endlessly. Either Long can consider adjusting the examples where he has previously discussed constitutive means along the lines of what you suggest, if that is all it takes to fix things. Or, in the case where Proposed Changes #2 may be needed, then we can possibly agree that the revisions are becoming extensive enough that Professor Long would need to assent to these further changes you or someone may propose in the presentation of his theory.

In short, my original critique was of Long’s theory as it was presented in the passage in question. At this point, I would prefer not to debate a follower’s or supporter’s revised version of Long’s theory.

Here is your counter-example:

“A better example would be:

Playing this chord here is an internal means to giving an accurate rendition of the Moonlight Sonata because it makes no sense to say “I wish I could give an accurate rendition of the Moonlight Sonata without having to play this chord here.” Therefore this chord here is a constitutive means of giving an accurate rendition of the Moonlight Sonata.”

In your counter-example, the phrase “this chord here” could have at least two possible meanings. It could mean “this chord here” in the sense of “this chord I am about to play.” Or it could mean “this chord here” in the sense of “this chord I now play,” where we conceive that I no longer have a choice in the matter, because the chord is in fact being played now.

If the phrase this chord here is given a precise or rigorous meaning or definition, this may lead to or necessitate further clarifications. For example, if the meaning of “this chord here” is this chord now played, and A has no choice in the matter, then the possibility exists that the theory envisions a series of such played chords where A has no choice in the matter, so that the Moonlight Sonata somehow “plays itself” through A without A’s choice.

If this is deemed an inadequate conception of things, then the idea of choice might have to be introduced into the theory or its presentation and examples. Doing so might complicate Long’s theory.

If this chord here means “this chord I am about to play,” then A has the option of playing it or not. This might complicate Long’s theory.

Here are a few other challenges that Long’s theory of constitutive means and/or the way Long has presented his theory may be confronted with:

Since the future is uncertain, no one knows whether person A will play all the rest of the notes of the Moonlight Sonata after playing this chord here.

What song A plays depends on the notes he plays after this chord here. He may play the first 47 chords of the Moonlight Sonata, and decide that the 48th chord—this chord here—will be the first chord of a new song.

Thus it is not necessarily true that: “Therefore this chord here is a constitutive means of giving an accurate rendition of the Moonlight Sonata.” This chord here might be the beginning of a new song. We don’t know for sure. Because in saying this chord here, we imply that we are in the present. But whether this chord is the middle chord of the Moonlight Sonata or the first chord of another song we won’t know until later.

Person A: I wish I could give an accurate rendition of the Moonlight Sonata without having to play this chord here.

Person A expresses his/her wish (desire, hope, prayer, etc.) that he/she didn’t have to play that chord there in order to render the Moonlight Sonata accurately. Person A may or may not get his/her wish (wishing may seem to others an improbable way for A to get what A wants). But A’s wishing it none the less makes perfect sense.

Person A: I wish I could give an accurate rendition of the Moonlight Sonata without having to play this chord here. I’ll have B play it for me, and my rendition will still be accurate.

I listed the above challenges in what you might consider a descending order of importance. You might believe that #3 is trivial. Still, it may require a small adjustment or change to the theory or its presentation. But such small adjustments add up, and after a number of them are made, they may add up to a substantial change in the theory or its presentation. So it is possible that even small and seemingly unimportant challenges need to at least be considered. Also, we don’t know for sure which issues, if any, Long may consider the most significant for his theory. We can’t assume that every theoretical issue we consider trivial will be considered so by Long.

.

22

Wombatron:

I numbered your replies 1-3 above.

  1. OK Then consider this question:

In Long’s theory, can this chord be the first chord of the Moonlight Sonata, and thus both constitutive of and instrumental to playing the Moonlight Sonata?

If not, why not?

  1. I disagree with you here. I think if you try to substitute the phrase “this chord here” for “all these notes played in just this sequence” Long’s example fails. The test Long provides for whether something is or is not a constitutive means occurs after the words “One test for the difference…” in Long’s example. So we have to take “this chord here” and plug it in after the words “One test for the difference…” in Long’s example. I don’t think his example works if we do this.

To understand Long’s theory we need some examples to go by, which I assume is the reason he provides examples. If the proposed substitution doesn’t work, then we need new examples.

  1. No. I’m trying to illustrate the idea that if we substitute “this chord here” for “all these notes played in just this sequence” in Long’s example, as I wrote above in #2, then Long’s example doesn’t work. Since my illustration may be flawed or confusing, then we can just revert back to Long’s own example.
23

Here it appears that the context Long is using in regard to “playing all these notes” is a context where the term “all these notes” is equivalent to the term “Moonlight Sonata”. “All these notes” is just another term for “Moonlight Sonata” in the example above. Long says “it doesn’t make sense” to attempt or wish to have the “Moonlight Sonata” without “all these notes”.

Here Long introduced the term “logically separable”. This time it appears that the context Long is using is a context where the term “fame, fortune, and the love of women” is NOT equivalent to the term “Presidential Address”. Long says “it makes sense” to attempt or wish to have “fame, fortune, and the love of women” without “Presidential Addresses”.

The only difference between the 2 examples is the context that Long puts them in. One “makes sense” and consists of 2 “logically separable” terms, while the other “doesn’t make sense” and consists of 2 equivalent terms.

Long’s goal appears to be an attempt to make a categorization of 2 types of means by showing that there are 2 fundamental relationships between means and ends which he calls internal means and external means. Long is essentially saying that an internal means is a means that is equivalent to its end, and an external means is a means that is non-equivalent to its end. However, means and end are never equivalent. Means are employed in order to attain ends. If the means is equivalent to the end (internal means), then no action would be necessary, we would be talking outside the realm of praxeological reasoning. Therefore, all that is left is external means. Which is essentially just another word created by Long for the traditional concept of just “means”.

24

The following questions are put to Roderick Long’s conception of constitutive means as described in his working paper “Wittgenstein, Austrian Economics, and the Logic of Action.”

http://mises.org/journals/scholar/long.pdf

The passage in question appears on page 127 in which Long explains the concept of constitutive means.

These questions are a follow-up to the short critique of constitutive means as that concept was presented by Long in his essay “Why Does Justice Have Good Consequences?” (posted above: A Short Critique of Long's Theory of Constitutive Means )

I will insert brief comments between the specific ideas Long conveys. Underlines are mine.

“The notion of a constitutive means is helpful here. Suppose one of my aims is to own a Rembrandt painting…..I just like having a real Rembrandt hanging on my wall.”

Long begins by supposing his aim as owning or having a real Rembrandt hanging on his wall.

“So I purchase a forgery, mistakenly believing I’m getting the real thing. Now I purchase this physical object for the sake of owning a Rembrandt;…”

In order to own or have a real Rembrandt hanging on his wall, Long will purchase a Rembrandt, not knowing that this particular Rembrandt is a forgery. Perhaps Long will take a bus to the art gallery where the forgery is offered for sale, negotiate a price, and write a personal check. Then, not wanting to risk damaging the forgery (which Long believes is a real Rembrandt), perhaps Long will take a taxi back to his apartment and then hang the forgery on the wall. The purchasing and the owning of the forged Rembrandt (or at least the having it on the wall) seem to be nonidentical, if this conception of things is what Long is intending in his example.

The idea of getting has been introduced, but is not significant at this point, since in this context, Long is still giving a somewhat typical account of “purchasing for the sake of owning or having on the wall.”

(A more typical means-type phrasing might be: “purchasing in order to…” or “purchasing so that…”)

It is not clear why the term “forgery” is replaced by the term “physical object” in the second sentence.

“…so acquiring the physical object is, in a sense, a means to acquiring the Rembrandt.”

“Forgery” is now permanently replaced by “physical object,” but there is no explanation why this change has been made. May we still consider “forgery” and “physical object” identical in Long’s example? May we substitute “forgery” whenever Long uses “physical object” in his example, and still have an example that describes constitutive means in a way acceptable to Long? If not, why not?

Long introduces a new concept, the concept of “acquiring.” But is acquiring purchasing, or is acquiring having on one’s wall? This is not clear in Long’s example. Nor is it clear why he changes terminology. Why doesn’t Long continue on with the ideas of “purchasing” and “having on his wall”? If Long were to carry through the original concepts of purchasing and having a real Rembrandt hanging on his wall, rather than switching to the idea of acquiring, would we still have an example that describes constitutive means in a way acceptable to Long? If not, why not?

What is clear is that now, in the phrase in question, “acquiring” is identical to “acquiring,” whereas earlier in Long’s example, purchasing and owning (in the sense of having on one’s wall) seemed to be nonidentical.

“Yet clearly I’m not regarding my acquisition of the object as an external or instrumental means to acquiring a Rembrandt;…”

It still is unclear how acquiring a Rembrandt relates to purchasing a Rembrandt on the one hand, and having a Rembrandt on one’s wall on the other hand. Long has not explained the switch or whether the switch is necessary for his example to work. Will a substitution of the original terms for the new term result in an equivalent demonstration of constitutive means?

Obviously, the acquisition of something is identical to acquiring something. But why has Long changed from the original ideas of purchasing, and having on his wall?

And why has the forgery become an object or physical object? May we substitute these terms as equivalents in Longs example? If not, why not?

“…rather, I believe (wrongly) that getting this object just counts as getting a Rembrandt, and so is a constitutive rather than an instrumental means to my goal.”

In this passage, Long reintroduces the idea of “getting” and uses it twice. Here, “getting,” like the idea of “acquiring,” is indeterminate in its relationship to purchasing on the one hand, and having on one’s wall on the other. Is getting purchasing? Is getting having on one’s wall? Is getting a combination of both, or is it maybe something else entirely? Long does not indicate or explain the change in terminology.

As with acquiring in the previous two passages, Long is now using two terms of equivalence: getting is identical to getting.

The original term forgery is still described as an “object.” In Long’s example, may we write equivalently:

“…rather, I believe (wrongly) that getting this forgery just counts as getting a Rembrandt,…”?

If we may, then may we also write “forgery” in the other instances where “object” or “physical object” is written?

Or if this does not convey the concept of constitutive means accurately, why is it necessary to change the concept “forgery” to “object”?

I assume that both Long and his followers would consider the above questions reasonable questions whose answers would result in a more clear understanding of the concept of constitutive means.

25

Adam, that was a very acute analysis. Long does seem to change his terms as he goes along, even from sentence to sentence.

You have astutely pointed out the ambiguity in the two instances of getting here: Does Long mean “purchasing” in the first and/or second instance? Does he mean “having on my wall” in the first and/or second instance?

However, I’d like to point out another possible problem with his proposition as phrased here.

Long’s goal is clearly in the form of a state/condition: “having a real Rembrandt hanging on [his] wall.”

Here Long clearly sets up the action, to “purchase a forgery” / “purchase this physical object,” as a means of some type. Long’s means here is clearly in the form of an action.

Now an action cannot be part of, or an element of, a state/condition in a rigorous sense. That would be a formal mismatch. However, later Long switches from “purchasing” to “getting,” and from “having on [his] wall” to simply “getting.” This allows him to sidestep the formal mismatch, so that “getting” can indeed be part of “getting.” But I think it’s clear from this vantage point that all he’s done is blur the lines.

Proof by fuzzy wording strikes again?

26

you’re a good case study of a proofreader

27

Wade, who posts here on the Mises forum, did some great research on the concept of constitutive means. The following is based on what his research has uncovered so far.

As far as we can tell, the original concept of constitutive means, at least in the context of Aristotelian eudaimonia, was introduced by the Aristotelian scholar L. H. G. Greenwood, in the introduction to his translation of Aristotle’s Nicomachean Ethics, Book 6 (Cambridge University Press, 1909).

I intend to post an analysis of the concept as it was originally introduced by Greenwood in 1909. The current post focuses on three important passages in Greenwood’s original presentation of this concept. On pages 46 and 48, Greenwood writes:

“But in considering the doctrine of the final end and of the means thereto, a new question arises, the importance of which does not seem to have been grasped by Aristotle himself; and so it is not easy to tell how he answers or would have answered to it. A thing may be a means to an end in either of two senses, as a component of it, or as wholly external to it.”

“Here then is a recognition of the distinction between component and external means: it is not recognized as a general principle, but as embodied in certain particular instances, which was as far perhaps as Aristotle ever succeeded in thinking it out.”

“I have attempted to indicate my own view, that the two notions of means are really combined in the sixth book’s definition of [Greek word here meaning: wisdom?]—or as it would be truer to say—that they were never properly distinguished from each other, but were both confusedly taken into account, artificially unified by their possession of a common name.”

There are at least two important aspects of these passages, written in the context of Greenwood’s original presentation of the idea of constitutive means. First, the conceptual distinction Greenwood introduces is, according to Greenwood, not grasped by Aristotle, nor is it recognized as a general principle by Aristotle. What Greenwood expounds as the concept that is today constitutive means, is something Greenwood sees as embodied in certain particular instances of Aristotle’s writings. As Greenwood writes, the concept(s) under discussion is a concept Greenwood sees as embodied in particular instances (implied in certain passages), but these concrete embodiments were perhaps the extent to which Aristotle ever succeeded in thinking the concept(s) through. And in book six, according to Greenwood, Aristotle confusedly took the two notions of means into account, and artificially unified them.

We might interpret this as meaning that the explicit distinction between two senses of the term “means” is, in so far as the distinction is made explicit, not a distinction made by Aristotle, but one supplied by Greenwood. The idea here would be that certain passages or even entire sections in Aristotle’s text do not always convey clear meaning to the reader, and clear meaning can only be attained by abstracting from the actual text, the conceptual distinctions deemed likely or appropriate by the interpreter. An interpreter of the text could conceivably, as a general rule, leave the unclear passages or sections somewhat “indeterminate.” Or, the translator can approach unclear passages and sections of the text with an eye toward abstracting from these unclear parts what to him seem to be their conceptual implications.

The latter appears to be the case with respect to the idea of constitutive means. Greenwood is abstracting from Aristotle’s text what he considers the conceptual implications of Aristotle’s discussion of means. Greenwood is not reporting a distinction explicitly recognized by Aristotle. He is abstracting from Aristotle’s text a conceptual distinction that would, at least, render conceptually distinct and ostensibly clear, what in Aristotle’s text is muddled or confused, and that would, in the best case, provide or suggest also, a workable conception of two senses of the term means.

Of course, the consistency and ultimate clarity of these abstracted conceptions as they have been rendered by Greenwood is an entirely separate matter.

The main point is that in abstracting from Aristotle’s text, Greenwood takes something indeterminate, and so to speak “actualizes” it into something more defined or concrete. In Greenwood’s doing so, we can understand the concept of constitutive or “component” means as itself a means: a means utilized by Greenwood to transform indeterminate passages of text into passages with ostensibly clear conceptual meaning.

The second important aspect of Greenwood’s passages is his explicit recognition that the two senses of means in Aristotle are “artificially unified by their possession of a common name.” Greenwood obviously suggests that the two notions under discussion could (even should), in his estimation, each be given a separate name. This is important, because as we will see when we look more closely at various conceptions of constitutive means, there is a sense in which when we speak of something as identical with itself, or when we speak about a complex event or object and the way it is constituted by or identical to all its constituent parts, we cease to speak about “means” as entities of human purpose. When we speak of “means” in this way, it is possible that what is being described are not strictly speaking, means, but relationships, definitions, or conceptions of identity. The question is whether those things we are calling constitutive means are things of the nature: When in action object or event X appears, then components a and/or b and/or c must appear, when object or event X is defined, conceived, or envisioned as identical with components a and/or b and/or c.

The question about Aristotle’s conception of means, Greenwood implies, is whether the two distinct notions embodied in certain textual instances would be more accurately conveyed by using two distinct names.

28

The following brief discussion concerns L. H. G. Greenwood’s presentation of the concept of component means (what are now referred to as constitutive means) in the introduction to his translation of Aristotle’s Nicomachean Ethics Book Six (1909, Cambridge University Press). The passages in question occur on pages 46 and 47.

Upon introducing the distinction between component and external means, Greenwood immediately provides two examples. Here we will consider the first of these examples:

“A thing may be a means to an end in either of two senses, as component part of it, or as wholly external to it. To take a trivial example, fire and basin and cloth are means to a pudding in the latter sense, suet [fat] and flour and currants in the former.”

What Greenwood writes above is essentially the following, which we will refer to as #1:

With respect to a pudding, fire and basin and cloth are external means, while suet and flour and currants are component means.

By referring to the fire, basin, and cloth as external means on the one hand, and the suet, flour, and currants as component means on the other hand, what Greenwood seems to suggest is that in some sense, the suet, flour, and currants just are, or themselves constitute the pudding, while the fire, basin, and cloth are, to use a popular expression “instruments” of making the pudding.

The problem with this conception can be seen if we simply re-write Greenwood’s example as #2:

With respect to a pudding, suet and flour and currants are external means, while fire and basin and cloth are component means.

Here, we conceive or imagine that a person goes out looking for suet, flour, and currants “in order to” make a pudding, but the pudding just is, or only is, or is constituted, when the suet, flour, and currants—gathered previously and externally—are combined with the fire, basin, and cloth.

When we look at things this way, the suet, flour, and currants become, or seem to become, the external means, while the fire, basin, and cloth become, or seem to become, the component means.

But in Greenwood’s example it is implied that the kind of means a thing is, is a quality of the thing itself, and not a function of how the thing is considered by a subject.

He doesn’t write or conceive: suet, flour, and currants may be considered as either an external or a component means. Rather he writes and conceives: suet, flour, and currants are component means.

His intended meaning seems to be that the kind of means things are is a quality of the things themselves. But as our example illustrates, suet, flour, and currants can be considered external means to a pudding, or so it appears. If we grant the validity of the concept of component means, it seems it would be more accurate to say that suet, flour, and currants may be considered as either external means or as component means to a pudding.

29

The following brief discussion concerns L. H. G. Greenwood’s presentation of the concept of component means (what are now referred to as constitutive means) in the introduction to his translation of Aristotle’s Nicomachean Ethics Book Six (1909, Cambridge University Press). The passages in question occur on pages 46 and 47.

Upon introducing the distinction between component and external means, Greenwood immediately provides two examples. Here we will consider the second of these examples:

“A thing may be a means to an end in either of two senses, as component part of it, or as wholly external to it. …Or again, Happiness being considered as the end, the contemplation of beautiful pictures may be considered rightly or wrongly as a means to this end in the component sense, the going to picture galleries as a means to it in the external sense…”

In this example, with respect to his conception of component means, we might interpret Greenwood as meaning that the end of Happiness may be considered as “co-present” with the means ‘contemplation of beautiful pictures.’ If we interpret Greenwood this way, we conceive that the means ‘contemplation of beautiful pictures’ is not entirely identical with the end Happiness, but is one component of the end Happiness, such that when the means ‘contemplation of beautiful pictures’ is something a person “now does,” then as a co-present accompaniment to this, the end Happiness is something that in some measure or in some sense, the person “now has.”

In thinking of things this way, we view the component means which is ‘contemplating beautiful pictures’ as in principle, interchangeable and/or combinable with other conceivable component means to the end Happiness, such as “having friendship” or “having health” etc. A person may not be contemplating a beautiful picture, but he may have health, or a person may have health and also be contemplating a beautiful picture, etc. But in some sense, when the means ‘health’ is achieved, or when the means ‘contemplating beautiful pictures’ is done, then co-present with these means is the end Happiness or some measure or sense of the end Happiness. While the component means to the end Happiness may be interchangeable and combinable, and while they may each come and go, the end Happiness that occurs when they are present is “a-temporally” related to them. The end Happiness attendant to the several component means is attendant to the means as a co-present accompaniment, and not as a subsequent or causal effect of the means.

If the above description is an accurate portrayal of component means as this concept has been indicated by Greenwood, then it seems reasonable to suppose that a person can aim for a particular component means to the end Happiness (such as health, friendship, contemplating beautiful pictures, etc.) armed with the knowledge that if he is successful in achieving the means, he will be successful in achieving the end (in this case Happiness) since the end is a co-present accompaniment to the component means.

If a person can aim for a component means to an end, and thereby attain the end as a co-present accompaniment to those means, then it would appear that in some sense he can try to achieve the means “in order to” or “for the purpose of” achieving the end, even though the end is not temporally separate from the means. Or stated another way, though the means may not be “external” to the end, for the acting person, the means may still be “instrumental” to achieving it.

30

Excellent findings, Wade! And superb exegesis and analysis, Adam!

31

Wow, great work Wade and Adam!

In this example, Greenwood’s end is the physical object pudding. He does not explicitly say how pudding can be an “end,” which seems to usually mean “a state or condition that one aims for.” Rephrasing “pudding” as a state or condition, the end would be something like “having pudding,” “possessing pudding,” or maybe “having pudding ready for consumption.” (To better recognize this, it may be helpful to note that in common English, “I want pudding” is shorthand for, “I want the condition of possessing/eating pudding.” In the same way, it seems only reasonable to interpret, “My end is pudding” as shorthand for, “My end is the condition of possessing/eating pudding.”)

If the reader agrees, Greenwood needs to explain in what sense flour can be a component part of the state/condition of possessing/eating pudding. Flour is certainly a physical component of pudding, but that fact alone does not seem to get Greenwood where he would like to go.

Here, Greenwood’s example end is a state/condition: happiness. He gives “contemplation of beautiful pictures” as a possible component means to that end. However, again he does not say how the action “contemplation of beautiful pictures” can be a component part of the state/condition of “happiness.” The reader is left to guess.

Since above Greenwood does not claim that flour [component means] in any sense causes pudding [end], it seems we can likewise assume here he does not mean that contemplation of beautiful pictures [component means] in any sense causes happiness [end] - or else it would be an external means. So what other conceivable relations of “componentness” are there? Besides co-presence, which Adam covered, the only other relation that occurs to me is simply that “contemplation of beautiful pictures” is itself one type of happiness. But if Greenwood means this, a question immediately jumps out: How can an action be one type of a state/condition?

32

FYI

I would like to provide an update on my research in regard to the search for the original distinction between “instrumental” and “component” means. A couple of weeks ago Adam had an intuition that while Greenwood might have introduced the distinction in the context of Aristotelian scholarship, he may not have actually invented the idea. “…did Greenwood absorb this concept from general philosophy, and then apply it to Aristotelian scholarship?”

Realizing that this was entirely possibly I started searching again. I quickly realized that I was taking a top-down approach by starting with Long’s conception and working backwards through time. There are many sources out there that say Greenwood made the original distinction. I decided to take a bottom-up approach instead.

My new technique was to search for really basic concepts in philosophy that are closely related to these concepts. For instance, I started searching for sources on concepts like “parts of wholes”, “equivalence”, “non-equivalence”, etc. The first thing I came up with was the idea of Gestalt, or Gestalt Psychology. While I found sources that predated Greenwood, it was a dead end.

My latest breakthrough was found merely by looking up the term “means” on wikipedia. The article for means is actually an article on “instrumental value” which pointed me to an article on “instrinsic value” which pointed me an article on “value theory”. And in that article on value theory under the section titled “Intrinsic and Instrumental Value” it indicates that the idea of instrumental versus intrinsic value was first discussed by Plato in The Republic. Further research indicates that the distinction between intrinsic and instrumental value was actually introduced by Glaucon, the older brother of Plato, in the conversation he has with Socrates in the very first section of Book II of Plato’s Republic.

Here Glaucon introduces the idea of 3 types of ends: 1. goods “…we welcome for their own sakes…”, 2. goods “…which are desirable not only in themselves, but also for their results…”, and 3. goods “…one would choose…only for the sake of some reward or result which flows from them”

I will propose here that it appears that Glaucon was attempting to arrive at the concept of means by splitting the concept of ends as it was conceived in that time into categories to account for the a priori concept of means. That possibly only ends were conceived before this, or means were blended in with the concept of ends, and Glaucon was attempting to make a clear distinction between means and ends. The fact that there are 3 types of ends Glaucon introduced does not take away from this. If we take the 3 categories above and we align them with the modern concept of “means and ends” we get 1.end 2.ends and also means 3.means.

Whether or not the idea of intrinsic and instrumental value as it is presented here somehow transformed into the idea of internal and external means between 380 B.C. (Plato’s Republic) and 1909 (Greenwood’s introduction) remains to be found. It is possible, due to L.H.G. Greenwood’s introduction to Aristotle’s Nicomachean Ethics, that this was transformed into the idea of 2 types of means and 1 end, a possible “misinterpretation” of Glaucon’s 3 types of ends in Plato’s Republic. It is possible that Greenwood was familiar with Plato’s Republic and attempted to apply Glaucon’s 3 types of ends to Aristotle’s work using his interpretation of Glaucon’s 3 types of ends.

33

I made some revisions to my post from yesterday. Refer to the above post which I have revised.

Where I said “it was first discussed by Plato in the Republic” I meant to say “the idea of instrumental versus intrinsic value was first discussed by Plato in The Republic”

I also added to the last paragraph: “Whether or not the idea of intrinsic and instrumental value as it is presented here somehow transformed into the idea of internal and external means between 380 B.C. (Plato’s Republic) and 1909 (Greenwood’s introduction) remains to be found.”

34

Mises forum poster Wombatron has posted a re-statement of the notion of constitutive means employed by Long and others:

http://wombatron.wordpress.com/

In what follows I will explain why I believe that Long and his followers fail to provide an alternative means/ends framework to the one utilized by Mises, Menger, and Austrian Economics generally.

Here is Wombatron’s account of constitutive means:

“There are 2 different ways that an activity can be subordinated to an goal (the end). In the first case, the activity is purely an instrument or means to the end. In the second, the activity is a component or constitutive part of the end. For example, buying a tobacco pipe is just instrumental to the end of smoking a pipe; it is not itself the activity of pipe smoking. On the other hand, lighting the pipe is an integral part of smoking (although that isn’t all that smoking is, of course); one can’t pipe smoke without lighting it. Thus, buying a pipe is just a means to the end, while lighting it is part of the end.”

In this example, lighting the pipe is a constitutive means to smoking the pipe.

Here is Menger’s account of means and ends, with some preceding sentences for context:

“The purpose of the theoretical sciences is understanding of the real world, knowledge of it extending beyond immediate experience, and control of it. We understand phenomena by means of theories as we become aware of them in each concrete case merely as exemplifications of a general regularity. We attain a knowledge of phenomena extending beyond immediate experience by drawing conclusions, in the concrete case, from certain observed facts about other facts not immediately perceived. We do this on the basis of the laws of coexistence and of the succession of phenomena. We control the real world in that, on the basis of our theoretical knowledge, we set the conditions of a phenomenon which are within our control, and are able in such a way to produce the phenomenon itself.”

(Investigations Into the Method of the Social Sciences, NYU Press, 1985, p.55-56)(Italics added)

This is an account of the means/ends analysis employed by Menger and Mises.

If lighting a pipe is constitutive of smoking a pipe, then if through my efforts, I am successful in attaining or producing that which is lighting the pipe, then I will be successful in attaining that which is smoking the pipe, since lighting the pipe is “constitutive” of smoking the pipe.

The first thing to note is that we haven’t left the domain of Mengerian/Misesian analysis in conceiving that an individual may aim for that (A) which is necessarily connected to something else (B), as a means to produce or attain that something else (B).

That is, if I am to achieve that which is lighting the pipe, because lighting the pipe is constitutive of smoking the pipe, then lighting the pipe is still an “external” or “instrumental” means for me. Tonight I want to smoke a pipe. Lighting a pipe is constitutive of smoking a pipe. So I’ll aim for lighting a pipe tonight after dinner, and by means of so doing, if I am successful, I will have produced ‘smoking a pipe.’

Lighting a pipe is an instrumental means to smoking the pipe.

So the first point is that if A and B are not identical, but they come together, then I can aim for A as a means to produce B. This is the same conception of things advanced by Mises and Menger.

The second point is the question of whether lighting a pipe is always constitutive of smoking a pipe, or whether lighting a pipe may occur, while smoking a pipe does not occur?

We may agree that: lighting a pipe = lighting a pipe, Rand’s A = A.

But does: lighting a pipe = smoking a pipe? Does A = B ?

It seems obvious that lighting a pipe need not always be accompanied by smoking a pipe.

Brand Blanchard provides an important passage for students of philosophy:

“Consider how we reject an antecendent as the cause until it achieves approximate identity with the effect. A man gets malaria, and we say that he has been bitten by an anopheles mosquito. Of course there is nothing in such a bite, so far as we can see, to make malaria necessary. But then are we quite clear that the bite is the cause? It clearly happens at times without an ensuing malaria. The disease must be caused by something nearer to it in time and space. This would seem to be the pouring into the bloodstream, by means of the bite, of a mass of parasites called plasmodia. But we cannot stop here either. The mere presence of the parasites in the bloodstream is not the cause; they may be present while their host shows no sign of the disease. They must not only be present; they must attack the red blood corpuscles in the stream. But even this is not the proximate cause, for there is still a temporal interval between it and the appearance of the recognized forms of the disease. Following the attacks of the parasites, the blood corpuscles are systematically drained of their haemoglobin. Is this, then, the cause? It is natural enough to say so, but we might still intelligibly ask, Could not this happen and the disease still not happen? So we again move nearer to the effect. Since haemoglobin is the means by which oxygen is conveyed to the tissues, its disappearance means that these tissues are starved and cannot function. Here at last we have reached a condition which cannot occur without the occurence of the disease. But then this condition is the disease; this starvation of bodily tissues is the essential constitutive factor in it. As long as the series of changes presents us merely with state A followed unintelligibly by different state B, we continue to hunt for a cause that will bridge the gap in both necessity and time.”

(Reason and Analysis, Open Court, 1991, p.452)

If lighting the pipe is not absolutely always accompanied by smoking the pipe, this means that the event which is lighting the pipe is not exclusively a component of smoking the pipe, but may be a component of other things besides smoking the pipe. For example, lighting the pipe may be a component of a movie scene, or a component of a demonstration on pipe lighting, or a component of an unfulfilled intention to smoke a pipe, etc…

In short, lighting the pipe is not exclusively a component or constituent of smoking the pipe. Lighting the pipe may be a component or constituent of many other things.

In this context, I believe that professor Hoppe’s stated reason for disowning Rothbard’s natural law approach is instructive. Hoppe writes, regarding Rothbard’s natural law program, which Rothbard conceived as one arriving at an objective ethic:

"It has been a common quarrel with this position, even on the part of sympathetic readers, that the concept of human nature is far “too diffuse and varied to provide a determinate set of contents of natural law.” (Theory of Socialism and Capitalism, p.235)

This criticism of natural law ethics could be construed as applying equally to the concept of constitutive means. One might hold that since a particular object or event may be a constituent part of a large array of other objects and events, that it may be impossible to arrive at a determinate set of contents of constitutive means. That is, a determinate and concrete thing, such as lighting a pipe, can be part of a “diffuse and varied” number of other events, and need not be part of the event that is “smoking a pipe.”

These are two reasons why I believe the notion of constitutive means, when this notion is conceived as “contentual” rather than formal, ultimately fails.

I do not claim that the notion of constitutive means has no validity. I claim that this idea only has coherence and validity when it is conceived formally and without content, which is the argument Mises advanced in his lifetime of work.

For example, walking toward one location is “constitutive” of walking away from another location. But this is a formal construct, and doesn’t say anything about the content of what is walked toward and away from. The idea of constitutive means has validity when it is conceived formally and absent specific content. But then it is a useless construct for the program of objective ethics, a program that is trying to arrive at a rational justification for concrete values and conduct.

35

smoking a pipe at time t1 is the means to smoking the self-same pipe at time t2 and is also and end in itself in so far as the person has the end of enjoying smoking at t1 and enjoying smoking at t2

36

Hoppe says a lot of things in the course of presenting his theory of argumentation ethics that he completely fails to back up or explain. This is just one example. But I do agree that Rothbard’s account of rights is inadequate.

This is rather irrelevant. What matters for the purpose of illustrating the distinction between instrumental and constitutive means is whether one can really smoke a pipe (in the common sense) without lighting it. Similarly, one can play the same note or even short sequence of notes in both the Moonlight Sonata and some other song, but in order to play the Moonlight Sonata one must play a specific series of notes in a specific order and manner, etc. Another example would be the difference between buying a golf club in order to play golf (instrumental means) and putting on the green at the ninth hole (constitutive means). Sure you can put on the green without playing a full round of golf or even go play put-put, but I don’t know of anyone who can play a full round of golf without putting on the green, not even Tiger Woods is that good (but even if someone could, I don’t think it would matter because for someone putting on the green during a game of golf, putting on the green is a component part of playing a game of golf, a categorically different activity in relation to playing golf than buying a golf club).

Given these considerations, I don’t think you’ve shown that “constitutive means” is at best a purely formal construct.

37

An action cannot be an end. Only states or conditions can be ends. We don’t aim to attain actions, we aim to attain states or conditions. I will demonstrate this now.

When we speak of the ends “playing Moonlight Sonata,” “smoking a pipe,” “playing golf,” or other actions that take place over a period of time, we are speaking loosely. To clarify the ambiguity, one need only ask at what moment the end has been achieved. Let’s examine the three examples case by case:

  • Playing Moonlight Sonata: “When has the end been achieved?”

Answer 1: “At the finish of the song.” If the last chord is yet to be played, the end has not been achieved. If you stop me before I can strike the last chord, I will not have achieved my end of playing Moonlight Sonata, which means playing it all the way through. Here we can clearly see that the end is not the action, but the state/condition of having completed the action of playing Moonlight Sonata all the way through**.**

Answer 2: “It’s achieved many times, once each time a note played.” Therefore, I can play one note and I will have achieved my end. All right, but then the end is not the action, but the state/condition of having completed the action of playing the first note.

Answer 3: “It’s achieved little-by-little, continually through the song, but only fully achieved at the end.” All this implies is that to have constitutive means requires that we have constitutive ends as well, bringing us fully out of the ends-means analysis, so this possibility is out.

  • Smoking a tobacco pipe: "When has the end been achieved?

Answer 0: “Once I am in a position to immediately smoke the pipe.” Then the end is a state/condition, not an action.

The rest follows the pattern above, with the same analysis.

Answer 1: “Once the pipe is fully smoked.”

Answer 2: “It’s achieved many times, once each time a puff is taken.”

Answer 3: “It’s achieved little-by-little, continually through the smoking session, but only fully achieved at the end.”

  • Playing golf: "When has the end been achieved?

Same analysis as with the pipe example above.

Answer 0: “Once we are in a position to immediately begin the game.”

Answer 1: “At the end of the golf game.”

Answer 2: “It’s achieved many times, once each time the ball is struck (or upon the completion of whatever action we may choose).”

Answer 3: “It’s achieved little-by-little, continually through the game, but only fully achieved at the end.”

We can see, then, that in no case can an action be an end, so “constitutive means” where a means (necessarily an action) is also an end (which cannot be an action) is a contradiction in terms.

Instead of saying that our end is “playing Moonlight Sonata,” “smoking a pipe,” or “playing golf,” we might more accurately say our end is “having played Moonlight Sonata correctly all the way through” (for an amateur), “finding the time to sit down at my piano and play Moonlight Sonata” (for a professional), “having reached a state where I can immediately inhale tobacco smoke from a pipe,” or “having made all the necessary arrangements to begin a game of golf.”

Once again, in a now familiar pattern, we can see that a misbegotten notion born out of the semantic ambiguity inherent in language loses its meaning once the ambiguity is cleared up.

Note: One final ambiguity needs to be cleared up. Regarding, “It makes no sense to wish to play have played Moonlight Sonata without having played this note. Therefore [playing having played this note]* is a constitutive means to playing having played Moonlight Sonata, because the note is both a means to that end and an end itself.”

*Uh-oh, if we use “playing” we can’t be talking about an end (state/condition), but if we use “having played” we can’t be talking about a means (an action). By using the more fuzzy concept of “the note” (underlined), the equivocation slips by unnoticed.

I hope it’s clear from this example how easily language can trip up even very careful thinkers like Long. Words are blunt instruments.

38

Geoffrey:

I guess what I’m seeing with the concept of constitutive means is the following pattern:

  1. Constitutive means is the idea that something is identical with itself: The moonlight sonata is the moonlight sonata.

Or:

  1. Constitutive means is the idea that a some thing A is a component of another non-identical thing B.

I assume most of us will agree with #1. But this idea might not contribute meaningfully to the theory of virtue, eudaimonia, or flourishing ethics. Then it seems we are saying something like: virtue is virtue, eudaimonia is eudaimonia, and flourishing is flourishing.

Unless someone makes a contrary argument, to me it seems as though we are forced to conceive of a “means” as somehow non-identical with the end. And I believe this accords with most, if not all, accounts of the concept of constitutive means. In the history of the concept of constitutive means, as far as I’m aware, there is always non-identity supposed between the means and the end.

This leads us to #2, or at least something like #2.

And the argument I have been advancing, and which I believe has not been adequately addressed, is that—in the context of a discussion about constitutive means—when we view the means A as non-identical to the end B, then it seems A can be obtained or attained, without obtaining or attaining B. In other words, A is or can be a means to other things, and need not be a means only to B. Or said another way, a person can “use” or “utilize” thing A, for other purposes than attaining or obtaining thing B.

In the discussion about constitutive means, proponents answer this by pointing to #1. But #1 is just going back to a statement of identity.

I’ll show you what I mean by using the examples you’ve provided:

Example A:

“What matters for the purpose of illustrating the distinction between instrumental and constitutive means is whether one can really smoke a pipe (in the common sense) without lighting it.”

Yes, a person can smoke a pipe without lighting it. A person can find a lit pipe, pick it up, and smoke it. Or, a person can be handed a pipe that was lit by another, and smoke it. So the thing which is A (my smoking a pipe), does not necessarily include the thing which is B (my lighting a pipe).

Similarly, one can light a pipe without smoking it. Doing thing B need not necessarily be a part of doing thing A.

We can say that my lighting and smoking a pipe cannot be done without my lighting and smoking a pipe, but this goes back to #1.

Example B:

“in order to play the Moonlight Sonata (A) one must play a specific series of notes in a specific order and manner (B)” (A and B added)

But this seem to revert back to #1 above. If you want to play the Moonlight Sonata, then play the Moonlight Sonata.

It seems to be a statement of identity, since the Moonlight Sonata is defined just as playing these particular notes in this particular order.

See R. Long in “Economics and Its Ethical Assumptions”: “…the Moonlight Sonata just is those notes in that order.” (italics and underline added)

Let’s say there is a board game called “The Moonlight Sonata.” One plays this game by rolling a dice, and then moving pieces around on the board. Then one can play The Moonlight Sonata without playing any notes whatsoever. For your original statement to be accurate, we have to move A and B in your statement towards identity. We have to define one as the other.

So I don’t think this overcomes #1.

Example C:

“Sure you can put on the green without playing a full round of golf”

I agree with this. Additionally, one can put on the green without playing even part of a round of golf. For example, one can put on the green as part of a golf lesson. One can put on the green as part of a contest that simply utilizes this particular green for this particular contest. One can put on the green as part of a movie scene, etc… So putting on the green need not be a means of playing a round of golf at all.

Example D:

“but I don’t know of anyone who can play a full round of golf (A) without putting on the green (B)” (A and B added)

But this is now #1. We are now going to define golf as an activity that includes putting on the green. And then we are going to say that if one plays golf, one must putt on the green.

One can easily play an entire round of golf while only chipping on the green. Chipping and putting are not identical activities. And so to make the statement “but I don’t know of anyone who can play a full round of golf (A) without putting on the green (B)” accurate (in the sense intended), one is now going to have to change the statement so that A and B approach identity. We’re going to have to define A as B. Then we are back at #1.

I think it’s important to reiterate the context in which the idea of constitutive means is put forth. In contemporary libertarian social theory, the idea of constitutive means is advanced by the Randian/Rothbardian objective ethics school as a way of circumventing or surmounting the challenge posed by Austrian School and Misesian theoretical value-subjectivism. In the latter theory, things are given meaning by a “meaning giving” subject, the individual actor. Value-objectivism is an attempt to conceive meaning (value, standards of conduct, etc…) as existing “objectively,” i.e., independent of individual action (independent of the consciousness of the individual actor).

The idea of constitutive means as I understand it, is that some things A, “just constitute” (are just part of) some things B, in “reality”. That is, some things (constitutive means), constitute other things (are components of other things), not by virtue of how an actor utilizes or intends to utilize those things, but rather by virtue of “objective reality.” And for this reason, the program of objective ethics is a valid theory (accurate and non-contradictory), since there are constituents (A) of things (B) that are “objectively” ascertainable as being constituents of B, independently of a subject that utilizes or intends to utilize A for B.

So far, I’m not convinced that this has been demonstrated in an unambiguous way. I don’t think your examples satisfactorily address the apparent shortcomings of the notion of constitutive means.

I believe that Austrian School and Misesian theoretical value-subjectivism stands as a more accurate and more consistent framework for conceiving human action, whether that action be “economic” action (action in relation to the market economy), or “ethical” action (conduct directed by one person toward another).

39

Can you provide any sources to back this up? Just googling around, I’ve found some papers that use the concept of constitutive means, and they don’t treat it as a controversial issue. So I’m inclined to think you’re missing some distinction or something is wrong with your theory of action.

40

It’s Long’s theory of action we’re talking about, and unfortunately he doesn’t seem to give a clear definition of means and ends, so we are left to guess. He seems to generally appeal to common sense, so I do the same. Let me make my bolded claim in a different way:

By definition, “constitutive means” makes no sense unless an action can be an end. However, Long does not say how an action can be an end. I go on to demonstrate some problems that occur if we imagine that actions can be ends.

Stating it this way, I don’t think outside sources are relevant. The question is purely one of the coherency of the concept. Can an action be an end? The notion of constitutive means hinges on the answer being “Yes.” I make a case for “No.”

To refute me, I think you’d either have to find a flaw in my argument or explicate a way in which an action can be an end.