ISSA Proceedings 1998 – The Concept Of Argumentation In Peter Singer’s ‘Practical Ethics’
Peter Singer’s “Practical Ethics” is – at least in Germany – one of the philosophical books of the last decades having gained the biggest public attention. It is discussed rather controversly by people with most different accademic and social backrounds. But so far, it seems to me, there hasn’t been an elaboration from the perspective of argumentation theory. This is surprising as Singer explicitly conceives ethics in a way that “allows reason an important role in ethical decisions.”(PE 8)
I agree with Singer on this as far as the words used are concerned; but I am not sure, if we understand them in the same way. There are several related questions to answer that will help to understand, what it may mean to allow reason i.e. argumentation an important role in ethical decisions: How does Singer argue himself? What emerges thereby as his notion of argumentation? Are there alternatives? What are the effects of the different conceptions of argumentation on the notion of ethics?
As these questions mix very much I won’t be able to answer them separately one by one. Starting with the first I will touch the others in order to come up with a more or less round picture of the whole issue.
My paper has four sections. The first section extracts argumentative traits from Singer’s book. The second one introduces two concepts of ‘argumentation’. The third section will confront the argumentative traits with these concepts revealing differing evaluations. In the last section I will show relations between formal argumentative aspects of the “Practical Ethics” and material ethical ones.
1. Argumentative Traits in Singer’s “Practical Ethics”
There is one pivot in Singer’s ethical thinking. It is what he calls “the principle of equal consideration of interests” (ECI). The ECI, he asserts, is the adaequate expression of universalisability and a sound basis of equality. (PE 19) It formulates the ethical postulate not to be selfish and it incorporates – and by this is intended to be resistent against – the fact that men are individuals and differ as such.
Singer grounds this principle on a kind of utilitarianism enabling him to say: “The essence of the principle of equal consideration of interests is that we give equal weight in our moral deliberations to the like interests of all those affected by our actions.” (PE 19) Singer renounces arguing profoundly for his utilitarian position. He admits that his view “is not the only possible view of ethics” (PE 8) and he maintains that it “may be treated as no more than a statement of the assumptions” (PE 8) on which his elaborations are based.To confirm the ECI there are three major examples of argumentation presented using the ECI as a premis. The examples refer to racism, sexism and a fictional society enslaving those that score low on IQ tests. The structural core of the three arguments is merely identical. I will therefore present just the first one to show, what they are like:
P(1) Who does not give equal weight in his moral deliberations to the like interests of all those affected by his actions acts morally wrong.
P(2) Racists give more weight to the interests of members of one race than to the interests of members of another.
C Racists act morally wrong.
The major premis of this classical syllogism (barbara) states the ECI. The minor premis describes features of racists in terms of the ECI. The conclusion says lapidarly that you shouldn’t be a racist. There won’t be controversy that this argument is valid in the sense that if the premisses are okay the conclusion is okay as well.
The particularity of this syllogism is, though, that the conclusion is out of question. On the other hand the major premis is just an assumption. And as we recall this argument is presented in order to confirm the ECI viz. the major premis. Singer uses the large consensus regarding the evelness of racism to corroborate the ECI by showing that being used as as major premis of an ethical argument it leads to the desired result. To harden this corroboration he works with a couple of structural identical arguments employing the ECI as major premis as I already remarked.[i]
Then Singer turnes to the more problematic issues. Again he merely uses the same structure of argumentation namely the barbara syllogism with the ECI as major premis to establish the conclusions. But now the ECI is no longer treated as a pure assumption. In these cases it is rather employed to make a controversal conclusion plausible. The range of topics being treated in this manner is wide and I will pick one where the structural identity with the example above is particularly obvious. In other instances the argumentative structure is less pregnant in so far as Singer refines and differenciates the premisses and conclusions in various admirably subtle and sophisticated ways. But where he talks about equality for animals this isn’t neccessary and he frequently points at the analogy to the racism argument. Here is the core of the argumentative structure:
P(1) Who does not give equal weight in his moral deliberations to the like interests of all those affected by his actions acts morally wrong.
P(2) Speciesists give more weight to the interests of members of one species than to the interests of members of another.
C Speciesists act morally wrong.
Perhaps nobody wants to be called a speciesist anyway, but Singer makes clear what it amounts to avoiding this lable: e.g. as citizens of modern urbanized societies we would have to more or less cease eating meat. (PE 45ff.) I will not discuss this attitude here, just point at the moral impact of this demand in Singer’s opinion: “If we do not change our dietary habits, how can we censure those slaveholders who would not change their own way of living?” (PE 56) This is apt to make some of us gourmets blush.
With this admittedly not very theoretical remark I will leave the exposition of some argumentative traits in Singer’s book. I will return later in section 3 to the presented examples after going into some more general aspects of the matter.
2. Two Notions of Argumentation
This section will roughly sketch two concepts of argumentation. The first notion is the one that is especially in english speaking countries common and seems almost selfevident. I will call it the “PPC concept” of argumentation (cf. Wohlrapp 1990: 232). According to this notion an argument consists in a set of premisses and a conclusion. The argumentative sequences extracted from Singer’s book that I have presented in the last section I have put in a form matching this conception.
The PPC concept is modeled obviously after a logical implication where the form of the antecedens and the form of the succedens govern a clearly defined relation between them. Accordingly the argumentative force is understood by the PPC concept as a kind of transfer of truth from the set of premisses to the conclusion warranted by the quasi logical form. Hence from the perspective of the PPC concept the best kind of an argument would be a deductive one where the truth of the premisses is transferred to the conclusion without any loss. But on the other hand informal logicians frequently have hinted at the relative poorness of this sort of argument due to its not coming up with any new information.
To conceive argumentation according to the PPC model means, to look for the structure of premisses and conclusions in a speech or text in order to grasp its argumentative content.
In contrast to this structure oriented concept I will now introduce a more action oriented approach (cf. Wohlrapp 1995 and 1998b, Mengel 1995: 135 161, Ch. 6). Instead of looking for a certain structure in speeches or texts we can ask for certain actions. There are three actions typical of argumentation: posing a thesis, substantiating and rejecting. The action of posing a thesis includes reformulating a once expressed thesis, in this way developping a follower thesis. The single actions attempting the substantiation of a thesis are called the giving of reasons, the single actions towards rejecting it objections.
On first sight we might relate the thesis of this concept to the conclusion of the PPC concept and the reasons to the premisses. This is not entirely wrong, but there are conceptional differences that shouldn’t be underestimated. We allready saw that the concept of posing a thesis has a potential dynamic component by including the possibility of generating a follower thesis. Furthermore, reasons in this model are not linked to the thesis in a logical or quasi logical way like the premisses to the conclusion in the PPC concept. Reasons are methodical steps on a way from a neccessarily assumed theoretical basis to the thesis. This basis contains established knowledge as well as current procedures and concepts. By “methodical steps” I understand actions that successively furnish conditions for the insight into the thesis. A very obvious difference to the PPC concept is the conceptualisation of objections (cf. Wohlrapp 1987). And the objections, explicitized or not, play an important role, because they are the actions that motivate substantiation. Without objections there seems to be hardly any need and no clear goal for argumentation.
A clou of this notion of argumentation is that it accounts for the possibility to integrate such objections: the dynamic concept of posing a thesis allows for reformulations of a previously expressed thesis incorporating the “wisdom” communicated by an objection. In the course of this generally the underlying theoretical basis is affected. It receives a sort of an “update”.
We see that there is not only a movement in one direction – like the transfer of truth in the case of the PPC concept. Besides the successive affording of conditions from the basis to the thesis there is also a supporting movement the other way round from the strengthened thesis back to a rearranged basis. As a consequence argumentation is not just understood as a probative procedure, but at the same time as an explorative action. This way good argumentation may very well be informative.
Harald Wohlrapp has called this aspect of forth and back in argumentation its “retroflexivity” (Wohlrapp 1990: 224[ii] and 1998a) Hence I will call this notion of argumentation the “retroflexivity concept” in contrast to the PPC concept. If we want to express the retroflexivity of argumentation in terms of the PPC concept, we might say that not only the premisses support the conclusion but also vice versa an accepted conclusion its premisses: Premisses and conclusion constitute a system of mutual
support. With these distinctions in mind we can now turn back to Singer’s argumentation.
3. Singer’s Argumentation and Retroflexivity
We remember that in Singer’s book we are confronted with two manners of argumentation nevertheless being structurally identical. They form the barbara syllogism with the ECI, the principle of equal consideration of interests, as their major premis. The difference of the two manners is that in one case, e.g. racism, the conclusion is commonly accepted while the ECI is treated as an assumption, in the other, e.g. speciesism, the ECI is used to support a controvers conclusion.
Reading Singer, who is very aware of himself being arguing, the impression is almost undenyable that he understands argumentation according to the prevailing PPC concept. So it is not surprising that he treats the part of his book, where he tries to establish the ECI and its utilitarian background, in a very tentative and cautious way using modest formulations etc.. This is the part using the first manner of argumentation (racism), which does not seem to match very well with the PPC concept, as there is no transfer of truth from the premisses to the conclusion.
Consequently Singer seems to view himself on firm ground when he turns to the more controverse issues and he uses his syllogism in the normal way. I already gave a taste of the rigidness of his dicta in these contexts. His argumentation here presents itself as an apodictical infering to conclusions that not everyone likes to accept despite the feeling that he is forced to.
So the PPC concept makes the first manner of argumentation trying to establish the ECI seem relatively weak, the second manner using the ECI seem rather strong.
On the background of the retroflexivity concept the whole issue appeares almost totally inverted. In this view the support of the ECI by means e.g. of the commonly shared disregard of racism is not peculiar at all. To demonstrate that it would be possible to use the ECI as a support of an affirmed attitude is doubtlessly a step on a way of insight into this principle.
But as an effect of the more dynamic conception of retroflexivity it is not as easily possible to separate the two manners of argumentation. Notably they are concerned with the same kind of topics and will therefore partly ground on the same theoretical basis. The retroflexivity concept of argumentation leads to taking the different PPC instances chiefly for components of a larger argumentation about the ECI. This means that on one hand indeed the instances with a very plausible conclusion strengthen the ECI, but on the other hand the instances with a problematic conclusion weaken it again. Every argumentative attempt of Singer’s, ending up with a controversial result, makes the ECI with its utilitarian background less trustworthy.
From this point of view one would expect Singer to become more cautious with his attitude and formulations in this part of the argumentation, but – as we saw – the opposite tendency is to be noticed. Not the successive probation of the ECI makes it seem stronger and stronger, but only constant mention and use of it. Frequent repetition of an opinion is a very old rhetorical device – just think of Cato’s famous “cetero censeo…” – but it is not regarded as a very noble argumentative means. In our examples, i.e. racism and speciesism, the mere rhetorical aspect receives support by Singer’s coining the expression “speciesism” in analogy to the expressions “racism” and “sexism”. This way the pejorative connotation of the latter will tend to be transfered to the former, effecting the evaluation of the matter in case.
This maneuver with expressions is probably intended, but I don’t believe that Singer is aware of the repetition effect. I rather suppose, his view is too much prestructured by the PPC concept. So I won’t blame him for playing bad rhetorical tricks on us. But the whole case shows that an inaedequate notion of argumentation is not just an argumentation theoretical flaw, but it may be apt to misrepresent the very topic of argumentation as well.[iii]
4. Argumentation in Ethics: Mementos
In this last section, devoted to relations between ethical issues and their argumentative treatment, I will become more tentative, because this is not the place and there is not enough space for an exaustive elaboration of this topic. But I feel obliged to at least give hints and perspectives for further reasoning about these matters. The discovery of the considerable doubtfullness of the ECI by looking through the glasses of the retroflexivity concept of argumentation should turn our attention back to the foundation of the ECI. We have to ask if there are further objections to it besides the possibility to generate questionable ethical demands by means of the ECI.
As far as I see, there are two major targets for objections in the formulation of the principle:
1. the assignment of interests is a problem,
2. it is not absolutely clear, what it is to “give equal weight” to “like interests”.
Regarding the first problem I can’t help viewing the notion of interests as a very private, subjective category. My interests are first of all only accessible to me myself, yours only to every single one of you yourselves. So my assignment of interests to you is for a rule dependent upon your conveyance of them to me. If we knew each other very well, I might be fit for good guesses in this respect or if I took you as a group and not every single, individual one of you. The two sources for the quality of these interest guessing are our communication on one hand, “if we knew each other very well”, and me recognising myself in you on the other, not taking you individually, but as representing an abstract self analogous to me.
To render possible the intended wide range of the ECI Singer cannot just rely on these two possibilities. Instead he objectivates the category of interests in a substancially naturalistic manner. This way he is able to extend the unmetaphorical use of the word interest without problems to unconcious men in very different life situations
from his own as well as to animals.
Nevertheless the naturalisation of a concept like “interest” raises problems. Friedrich Kambartel has introduced the term “grammatical threshold” (grammatische Schwelle) into the discourse about changing concepts in such a way (Kambartel 1989: 71). He takes up Wittgensteins use of the expression “grammatical” and states that such grammatical thresholds lie in between the lingual means of two entirely different fields of language use. The
naturalisation of the concept of interest would mean the nivellation of a grammatical threshold in this sense with possible grave ethical consequences. On this way we might end up considering the “interests” of tamagochis.
I cannot elaborate this any further here for the stated reasons. I just wanted to show the kind of objections Singer would have to cope with in order to establish the ECI, confirming it as much as to allow it to bear all the questionable ethical demands.
Probably even more problematic are the objections that focus on the idea of giving “equal weight” to “like interests”. As I have already remarked, Singer holds the ECI to be capable of showing “why the most blatant forms of racism, like that of the Nazis, are wrong.”(PE 20). But why shouldn’t the Nazis have argued: What is the interest not to suffer of a few million jews in comparison to the interest of generations of mankind to lead a sublime life?
This illustration points to three interacting shortcomings:
1. Interests have to be criticisable. But as the ECI is conceptionalised as a principle, there is no basis for such a criticism within Singer’s system.
2. The quantifyability and summarisability of interests is questionable even though Singer claims that precision is not neccessary (PE 53). Here we eventually face again the naturalism issue.
3. To engage in the utilitarian calculation act – regardless wether you calculate interests or happiness or whatever – distracts from the qualities of actions as it presupposes a quantification.
All these possible targets of objections would not be grave, if the ECI were just a rule of thumb among others. But Singer introduces it as the ethical principle. As such it has to be pertinent for all the possible single instances of moral reasoning. To me it seems that the interpretations of the single situations being neccessary to allow this pertinence generally presuppose a considerable amount of moral judgement which can’t be backed by the ECI, too.
I suppose that a major motivation to conceive the ECI as a principle is the wish to make ethics accessible to argumentation. If you have principles you have secure premisses for e.g. syllogisms. But we have seen we are not forced to adopt a notion of argumentation being dependent on the existence of confirmed premisses. If we can assume some kind of a theoretical basis there is enough to start reasoning. Argumentation itself has the potential to extend, modify and confirm – or dismiss – this basis as far as neccessary by its explorative aspect. We’d better rely on this aspect in ethical reasoning instead of paving one way streets leading into blind alleys.
i. It is certainly possible to analyze Singer’s argumentation using a different form from the barbara syllogism. Singer’s frequent reference to the analogy of the speciesism argument – where the barbara structure is obvious – shows, though, that he himself is understanding his argument in this way. As I am concerned with his notion of argumentation I have to analyze accordingly.
ii. In this article Wohlrapp seems to restrict retroflexivity only to special cases of argumentation. Further research shows nevertheless that cases without retroflexivity seem to be rather exeptional. Cf. Mengel 1995: 196-199.
iii. I want to underline that I don’t argue against the PPC structure as an analytical tool – how could I, using it myself in this paper. It is the restricted notion of argumentation resulting from an more or less exclusive orientation from this structure that I want to stigmatize.
Kambartel, F. (1989). Zur grammatischen Unm÷glichkeit einer evolutionstheoretischen Erklõrung der humanen Welt. In: F. Kambartel, Philosophie der humanen Welt (pp. 61 78), Frankfurt a. M.: Suhrkamp.
Mengel, P. (1995). Analogien als Argumente. Frankfurt a. M.: Lang.
Singer, P. (PE). Practical Ethics. Cambridge: Cambridge University Press 1979.
Wohlrapp, H. (1987). Toulmin’s Theory and the Dynamics of Argumentation. In: F.H. van Eemeren et al. (Eds.), Argumentation: Perspectives and Approaches, Proceedings of the first ISSA Conference on Argumentation 1986 (pp. 327 335), Dordrecht: Foris.
Wohlrapp, H. (1990). _ber nicht deduktive Argumente. In: P. Klein (Ed.), Praktische Logik (pp.217 235), G÷ttingen.
Wohlrapp, H. (1995). Die diskursive Tendenz. In: H. Wohlrapp (Ed.), Wege der Argumentationsforschung (pp. 395 415),
Stuttgart Bad Cannstadt: frommann holzboog.
Wohlrapp, H. (1998a). A New Light on Non-Deductive Argumentation Schemes. In: Argumentation 3, 1998.
Wohlrapp, H. (1998b). Frames and Positions: On Grounding Argumentation Theory Pragmatically. In: Proceedings of the 10th SCA/AFA Conference at Alta, forthcoming.