Critical Study INFORMAL LOGIC XV.3, Fall 1993 Slippery Slope Arguments by Douglas Walton WIBREN VAN DER BURG University of Utrecht Walton, Douglas. (1992). Slippery Slope Arguments. Oxford: Clarendon. Pp. xiii, 1-296. CDN $82.50. A pragma-dialectical approach to slippery slope arguments In this book Walton argues that some- times the slippery slope argument "can be used correctly as a reasonable type of argu- mentation to shift a burden of proof in a critical discussion, while in other cases it is used incorrectly". I But this incorrect use is only seldom a fallacy. What makes this book interesting is not this general thesis, but the wealth of detailed analyses of slip- pery slope arguments. The book begins with a general definition. A slippery slope argument is a kind of ar- gument that warns you if you take a first step, you will find yourself involved in a sticky sequence of consequences from which you will be unable to extricate your- self, and eventually you will wind up speeding faster and faster towards some disastrous consequences. 2 (I will refer to the first step as A, to the in- termediate steps as AI, A2, etc. and to the undesirable end result as B.) Thus the slip- pery slope argument is a form of practical reasoning, usually aimed at convincing someone that a particular action should not be chosen. Walton advocates a 'pragma-dialectical' approach to the slippery slope argument. This approach (as a general approach presented in various other books and articles by Walton) focuses on how an argument is used in actual dialogues, taking into account the specific context of dialogue in which it is used. Instead of abstract theorizing, a case study approach would be the best meth- od. The pragma-dialectical approach leads to a novel analysis of fallacies. According to Walton, a fallacy is a basically reasona- ble type of argument that has been serious- ly misused in a concrete instance, violating the rules of the dialogue. 3 The slippery slope argument is basically reasonable, and in most cases is not so badly wrong that it should be categorized as fallacious. As a slippery slope presents an argu- ment about anticipated future consequen- ces, it is only probabilistic. How probable these consequences are depends on the facts of the concrete case. Therefore Walton suggests we should assess the de- tails of the case before we decide whether a slippery slope argument is convincing. This judgment is always defeasible: new evidence may call for revision of a prior probability judgment or may point to out- weighing beneficial results. The slippery slope argument is nevertheless often effective in throwing enough doubt on a proposed action to shift the burden of proof to the respondent. Walton summarizes his basic approach as follows: Slippery slope arguments are characteris- tically: (1) uses of practical reasoning; (2) used in a context of dialogue, meaning 222 Wibren van der Burg that they are bilateral, involving a proponent and respondent; (3) negative arguments from consequences; (4) defeasible; (5) of varying degrees of strength or weakness, but rarely outright fallacious; (6) often effective in shifting the burden of proof.4 Walton distinguishes four main types of the slippery slope arguments. For each of these he presents a detailed analysis of the argumentation schemes, strategies, and identifies critical questions. Walton con- cludes that only when the proponent of the slippery slope argument is pushing too hard or uses an overly strong formulation, may we properly call it a fallacy. 1. The first type of slippery slope argu- ment is the sorites. The most famous ex- ample of it in the context of practical reasoning is the abortion case: one more day cannot make an essential difference for the moral status of a foetus as a human person, therefore how can there be a non- arbitrary point between conception and birth? (Interesting is that Walton shows that this can be the basis for a slippery slope argument both pro and contra abor- tion.) Walton argues that the sorites is not a fallacy, but a paradox which arises from the vagueness of language. This vagueness is especially problematic when scientific or technical development creates new situ- ations that are not covered by traditional terms, like 'death', 'life', and 'person'. The sorites argument depends on the use of key terms that are clear in some are- as of application (the white area) and un- clear in others (the grey area). But-and here Walton makes an important point-the problem is not just the fact that these two areas exist, but that the transition from the clear area to the grey area is itself grey. The indeterminacy of the transition leaves the respondent to a sorites argument with- out a clear point of defense. And this is ex- actly the reason why the sorites argument may be a valuable warning: if there is no clear line to draw, the transition from white through grey to black seems plausible. This danger may be enough to shift the burden of proof. The best strategy for the respondent is usually to deny the applica- bility of consistency, even to closely similar cases, on the grounds that the key term is vague. 5 In fact, any arbitrary line in the grey area will do. Another way is to propose a more precise definition of the vague teon. 2. The second type is the causal slip- pery slope argument in which there is a causal chain of events leading to an unde- sirable outcome. This is the least convinc- ing chapter in the book, because the treatment of social processes is superficial. When saying that actions A and B are both events in a causal chain, we need to know that they are not merely symbols or symp- toms of a broader process and that B is not merely a side-effect of a generally positive social process. The growing acceptance of the doctrine of informed consent (A) may plausibly be regarded as a symptom and not as a cause of a broader process in which autonomy is increasingly respected, which process perhaps in the future may result in a growing acceptance of infanticide (B). But how do we distinguish whether the acceptance of abortion (A') must be re- garded as a symptom as well or as a cause of B?6 To answer this type of question, we need a theory of social causation. Walton's analysis remains within the context of dia- logue. Consequently, it is to be faulted for not addressing this central problem. 3. The third type of slippery slope ar- gument is the precedent argument, in which allowing an exception to a rule is seen as setting a dangerous precedent. Many ethicists do not treat this as a separate type.! In fact, the precedent argu- ment and the sorites argument may be seen as two sides of the same coin: the former saying that once a precedent is set we can no longer draw a line, the latter arguing that the reason for this is that we do not have principled and non-arbitrary reasons because the vague terms used in the argument to justify A, also apply to AI... through B. In view of this difference of opinion, it is regrettable that Walton no- where argues why we should see sorites and precedent as separate. Walton distinguishes various subtypes of the precedent slippery slope. 8 The sim- ple dangerous precedent argument is the basic type, arguing that the first step will lead us to horrible results. The arbitrary re- sults argument holds that once we set the precedent, there will no longer be a non- arbitrary and clear criterion to draw a new line. The feasibility precedent type argues that, though each of the following conse- quences AI, A2 and so on through B, in it- self is acceptable, allowing all these consequences together is unacceptable. An example of this type is the following one: "allowing a few immigrants would not be a problem, but our country cannot absorb all that may want to immigrate, therefore we need to stick to strict immigration rules". Walton's fourth SUbtype is the argument from added authority: "allowing an initial precedent would grant additional authority to an institution or decision-maker, which would lead to a sequence of bad results, in- cluding some ultimate, horrible outcome".9 The precedent slippery slope argument builds on the ideal of consistency (or on the closely related ideal of universalizability). A logical inconsistency within a well- defined set of propositions held by the pro- ponent is very rare. Therefore, Walton ar- gues, we should focus on 'circumstantial' inconsistency. Circumstantial inconsistency "depends on the presumption that certain personal actions or other personal circum- stances suggest or make plausible that there may be a logical inconsistency inherent in an arguer's position."10 The pragma- dialectical approach has an advantage here: in the context of a dialogue circumstantial inconsistencies can be explored more easily because we may articulate and then openly discuss this kind of implicit inconsistency. 4. The last type is the most interesting one: the full slippery slope argument. It combines all three of the simple slippery Review of Douglas Walton 223 slope arguments in one complex structure, together with a fourth element, an appeal to a social climate of public opinion. Walton rightly remarks that in many actual debates it is not clear whether the supposed links in the sequences are meant to be causal, prec- edents, or sorites connections of similar cases. 11 Some of the steps may be of the sorites type, others of the causal type, and again others combine various types. But an important factor in making the slippery slope argument plausible is usually the so- cial climate of acceptance that propels the movement along each of the steps down the slope. This reference to (predictions of) public opinion makes it a difficult argument both to attack and to defend. In the last part, Walton tries to develop a more general theoretical framework. The slippery slope is a negative variant of the argument from gradualism. Using the pro- posed course of action A as an initial premise, a presumptive modus ponens is applied repeatedly until a horrible result B is reached. Then going backwards by a se- ries of presumptive modus tollens, it is shown that the initial step must be rejected to avoid this horrible result. Walton gives many practical sugges- tions on how to use and counter the argu- ment. He distinguishes six basic tactics to counter a slippery slope argument. The first is that we can claim that the negative consequences won't really fol1ow; if we use this tactic we should usually attack the weakest links and/or the earliest links in the chain. Secondly, we may cite the un- certainty of the future. Modifying the goal to eliminate the negative consequences is a third tactic and stressing positive conse- quences that outweigh the negative conse- quences is a fourth. We may try to choose some alternative means of achieving the goal without the negative consequences. Finally, we may argue that not taking the action in question will have even worse consequences. The study of these tactics and subtactics are really a must for every- one who is practically or theoretically interested in the slippery slope argument. 224 Wibren van der Burg General critique Walton's book is an important contri- bution to the study of slippery slope argu- ments. He presents many interesting cases that are helpful in grasping the intricacies of the argument. Especially in some of the more strict analyses, he offers fresh in- sights into the structure of the slippery slope argument, in the way the argument may be used, and countered. This book will be a necessary starting-point for future studies of the subject. It is only a starting-point, however. There are too many problems left open and neglected. Moreover, the book would have benefited from more critical reflection on style and presentation. There are too many careless formulations, mistakes and impre- cise arguments. The exact relationships, similarities and differences between the slippery slope arguments and related argu- ments like the argument from gradualism or the ad hominem argument are often vague and obscure. Most irritating, howev- er, is the abundance of unnecessary repeti- tions. In particular, the theme that the slippery slope argument is weak but not fallacious is repeated over and over again. Had Walton been more self-critical, the book could have been significantly shorter and more convincing. A more important critique is that Walton's analysis is unconvincing on three major points. In the rest of this article I will focus on these three points of critique and tentatively suggest how we can come to a fuller understanding of the slippery slope. Definition and structure Walton's definition of the slippery slope argument is a standard one. Reflec- tion on some cases, however, shows the in- adequacy of this standard definition. The following case is presented as a causal slippery slope: "Suppose we are consider- ing whether to allow a manufacturer of hy- draulic fluid to dump millions of gallons of PCB-contaminated wastes into a small stream. Someone who opposes such dumping might argue that the PCBs will run from that stream into a downstream river, will accumulate in fish, will pollute our drinking water with a known cancer- causing agent, and will eventually result in pollution of rivers, killing of wildlife, and severe hazard to humans who use the water downstream." l2 Under the definition this qualifies as a slippery slope argument, but nevertheless something is missing. I would suggest that it is essential for a slippery slope that it is not merely a sequence of events, but a se- quence of actions. In this case there is a se- quence of events but not a sequence of actions: the only relevant action is to allow the manufacturer to dump wastes. The fur- ther sequence of events is just a 'natural' causal chain which in the long run has dis- astrous effects. This is merely a negative argument from long-term consequences. If I warn you not to throw a snowball to avoid an avalanche, this is not a slippery slope argument, but simply a warning of disastrous results because of one action. But we need more than a mere se- quence of actions. Walton considers the ar- gument from added responsibility as a subclass of a precedent slippery slope ar- gument. He defines the argument as "al- lowing an initial precedent would grant additional authority to an institution or de- cision-maker, which would lead to a se- quence of bad results, including some ultimate, horrible outcome."lJ He gives the following illustration: if one had argued prior to 1933 that if Hitler had gained power horrible things would happen. it would seem in historical retrospect that the argument would have had some merit." Of course this argument has merit, but it is not a slippery slope. If I advise my friend not to lend her car to Bob, because he drives carelessly and might have an acci- dent, there is only one relevant action in- volved: lending a car to Bob. Thus this is an argument of the type that we should not give a power or authority to X, because X is not the right person or institution to use it. It warns of a possible abuse of power. However, that is not a slippery slope argument. The Nazi example includes a hidden slippery slope, but the initial premise is not the granting of power to the Nazis (which was wrong for independent reasons). The first step on the slope is: not to protest against the first signs of oppression by the Nazis, against the first abuse of power. It is expressed in Martin Niemoller's famous quote: The Nazis first came for the communists, and I didn't speak up because I wasn't a communist. Then they came for the Jews, and I didn't speak up because I wasn't a Jew. Then they came for the trade union- ists, and I didn't speak up because I wasn't a trade unionist. Then they came for the Catholics, but I didn't speak up because I was a Protestant. Then they came for me, and by that time there was no one left to speak for me. If the argument from added responsibility is not a slippery slope argument, we need additional qualifications in the definition of slippery slope arguments. I would sug- gest two additional characteristics: these actions must be by the same actor or group or institution, and these actions must be similar in relevant aspects. If I make an of- fensive remark and someone else reacts by starting a fight that results in a vendetta, this is not going down a slippery slope (as it would be under Walton's definition) but a spiralling process of action and reaction: the characteristic "by the same actor" ex- cludes this. And if I allow the boy next door to play in my garden and while play- ing he accidentally sets fire to my house, this should not be regarded as a slippery slope either: we need the additional criteri- on that actions should be similar in relevant respects to avoid this. A related critique focuses on Walton's analysis of the composite nature of the slippery slope argument. Review of Douglas Walton 225 One party is warning a second party (by an argument from consequences) that some third party (a real or hypothetical proponent) will drag him along by an argument from gradualism, or by some gradual process, towards a conclusion that will be disastrous for him (the second party) .... A genuine slippery slope argument requires all three of these characteristic participants to be engaged in the roles attributed to them above. IS Walton's analysis is very unconvincing here. Firstly, the notion of third party agen- cy must be extremely broad to include na- ture and popular opinion as third parties. Secondly, not even a second party seems always necessary. Once I allow myself to rise at 7.40 a.m. instead of 7.30 a.m., be- cause of my lack of will-power, I will end up rising later and later; therefore I should not break my own rule. It is well conceiva- ble that some other person is trying to con- vince me to stay in bed longer; but this is not essential, let alone that there should be some third party. Walton here seems to turn things up- side down. Perhaps it is useful to analyze how an argument is used in the context of dialogue. Yet, we should not conclude from a two or three party structure of the analytic instrument that the reality (the ar- gument) itself has this structure. Slippery slope arguments are used monologically as well, and a pragma-dialectical approach should be able to account for that. Determining validity and plausibility Walton's goal is to prove that a slippery slope argument is often a weak argument but seldom so weak that it must be consid- ered a fallacy. In this proof he succeeds, but in a trivial sense. Once you redefine 'fallacy' in a pragma-dialectical way (see above, p,221), it is obvious that a slippery slope argument will infrequently be falla- cious. But more important, it seems to me that this approach to fallacies-at least in the case of the slippery slope argument-is for practical uses not very helpful. What does it add to our understanding of practical debates that we can not merely 226 Wibren van der Burg say that someone is arguing in an unrea- sonable way, but can qualify this as a falla- cy? Even if I succeed in proving my opponent's argument a fallacy, it still does not mean that J win the discussion. My op- ponent might simply reply by a more care- ful, non-fallacious formulation of the argument. The real question is when and why a slippery argument is valid and plausible. Structuring the slippery slope argument as a form of presumptive reasoning is not very helpful in this respect. It obscures the fact that the ground for this presumptive- ness is either based on analytical reasoning or on empirical probabilities, or on a com- bination of both. Either the slippery slope is a precedent or sorites one, and then the question is whether A, A 1 and so on up to B are relevantly similar and whether we can draw a reasonable and effective line somewhere between A and B.16 Or, it is a causal argument or a reference to the cli- mate of public opinion, and then it is a question of whether the suggested se- quence is probable. Whereas the superfi- cial structure may be analyzed in terms of presumptive reasoning, the deep structure is one of analytical andlor probabilistic reasoning. This critique, however, points to an in- teresting explanation of why the slippery slope argument is often more convincing in actual debates than is warranted. We treat it as a sequence of presumptive argu- ments, and as each individual step seems unobjectionable, we feel obliged to go along the slope. This tension between the superficial, dialogical structure and the deep structure is a central characteristic of the slippery slope argument to which Walton's approach blinds us. A full analysis of the slippery slope ar- gument should address the grounds on which it is held to be reasonable. This means dealing with normative problems about consistency and universalizability: why are these acceptable ideals and in what contexts are slippery slope arguments based upon them valid? And it means deal- ing with empirical questions: what factors are important to determine the probability of causal sequences? An analysis of the slippery slope argument neglects these questions at its own peril. Therefore, Wal- ton's pragma-dialectical approach should be supplemented by substantive studies in practical philosophy and social sciences. 17 The emotional appeal of the slippery slope argument Walton presupposes a model of reason- able dialogue as a collaborative social ac- tivity. He excludes reference to emotions or psychological factors in his explana- tions. The following passage is illustrative: Curiously then, the plausibility of a slippery slope argument often stems from an emotional pull to which a particular audience or respondent is susceptible. 18 To me this does not seem curious at all. The emotional appeal of the slippery slope argument explains why many ethicists are so suspicious about it. But instead of the obvious conclusion that we must find the causes for this emotional appeal, Walton excludes all emotional and psychological factors from his analysis. 19 Moreover, Walton seems to presuppose that it is possible to have a rational discus- sion free of emotions, in which the part- ners in the dialogue can find a common ground in non-controversial facts. My ex- perience with slippery slope arguments in the practice of biomedicine is exactly the contrary: they are often strongly dependent on controversial perceptions and construc- tions of reality. How you perceive reality and extrapolate trends and possibilities to the future is closely related to your basic attitudes and world views. Let me illustrate this point with one of the most controversial examples of a per- ceived slippery slope: euthanasia in the Netherlands.20 J have the impression that most physicians, lawyers, and ethicists in the US believe in something like the following story: In 1973 the Dutch set the first step on the slippery slope; they tolerated active volun- tary euthanasia on request in cases where death was near and where there was un- bearable suffering. But subsequently they abandoned each of these criteria by small steps. Now they are even discussing 'eutha- nasia' without request in cases of comatose patients, psychiatric patients and severely handicapped newborns. There seems to be no end to this sequence: we may expect them to go further down the slippery slope yet. On the other hand, most Dutch physicians, lawyers, and ethicists seem to perceive the Dutch history on euthanasia quite differ- ently, somewhat like the following story: In the late sixties, we began to realize that modem medical technology is not always beneficiaL Life is not always worth living and sometimes suffering is so unbearable or the quality of life so poor that prolongation of life is itself an eviL For the last thirty years, Dutch society as a whole has been involved in the process of this general dis- cussion on medicine and health care, in- cluding topics like medical decisions concerning the end of life. This broad and intense discussion has been long and diffi- CUlt, but gradually we have been moving towards some general agreement. The con- sensus started with the relatively easy cases: euthanasia in cases where there is a clear request and unbearable suffering and where the end of life is near. We went on to dis- cuss the more difficult cases and we are still struggling with them. Examples of the most. challenging cases are psychiatric patients who request euthanasia, comatose patients, and handicapped newborns. The process of discussion on these cases continues. 21 This perception of the Dutch story is not one of steps on a slippery slope, but that of a winding road. 22 For many years, the Dutch have been trying to convince their US colleagues of their-and what seems to me the correct-interpretation of the story, but usually in vain. 23 Here we have an interesting problem which seems characteristic of many slip- pery slope arguments. The same reality is perceived completely differently. If opinions differ so strongly about the interpretation Review of Douglas Walton 227 of a historical process, the differences will even be larger when discussing future devel- opments. For instance, consider the recent initiatives in Washington and California which would have allowed certain forms of euthanasia or physician-assisted suicide. If discussions about interpreting the Dutch situation have been in vain, how can we expect agreement on the assessment of the risks to follow these initiatives? To answer this type of question, we cannot exclude psychological and emotional factors. We need to address these factors directly, be- cause ultimately they seem to determine whether some person or group believes in the slippery slope or not. In the Dutch euthanasia example, these factors may be quite complex. One expla- nation is that many Americans simply con- demn every form of active euthanasia; every step will then clearly be perceived as a step down the slippery slope. A second explanation is that whether one perceives a development as a slippery slope largely de- pends on basic attitudes of trust in other persons and in society in general. In the US there seems to be much more distrust of doctors, lawyers, politicians, and fel- low-citizens (like family members) than in the Netherlands. The Dutch practice heavi- ly leans on trusting doctors, because legal control of medical euthanasia practice is extremely difficult. Doctors trust doctors, patients trust doctors, and the legal system entrusts doctors with these decisions. If someone with a basic attitude of distrust looks at this situation, he will see an ex- treme danger of abuse. A third explanation is that one implicitly always interprets a development in the light of familiar facts and values. In the Netherlands, there is an almost equal access to health care for eve- ryone and almost no one will have person- ally to pay extremely high hospital bills; euthanasia is usually performed in the con- text of a long-standing doctor-patient rela- tionship; and there has been a long, intense, and broad discussion on euthanasia. These facts are essential to understanding 228 Wibren van der Burg why the risk of a slippery slope is per- ceived as minimal in the Dutch society. If one lives in a society where the facts are different, one will more easily perceive the risk of a slippery slope. These explanations are very tentative. Other explanations may focus on the im- portance of personal experience, or on cul- tural values. But it seems that if we really want to understand the appeal of a slippery slope argument, we cannot avoid searching for this type of explanation. Even though Walton has made some progress in understanding the slippery slope argument, the major steps still have to be done. In the light of my critiques, I doubt whether a pragma-dialectical approach will be very useful in this ongoing search.24 Notes I Walton, o.c., 2. 2 Walton, o.c., I. The sequence is not necessary, but merely probable. 3 Walton, o.c., 22. 4 Walton, (J.e., 13. 5 Walton, o.c., 58. 6 W. van der Burg, 'The Slippery Slope Argument', Ethics 102 (1991): 42-65, repro in The Journal of Clinical Ethics 3(1992)4 256-292, esp. 264. 7 1. Glover, Causing Death and Saving Lives (Harmondsworth: Penguin, 1977) 16; T.L. Beauchamp and I.E Childress, Principles of Biomedical Ethics (New York: Oxford Univer- sity Press, 1989, Third edition), 139-141 both distinguish a logical and an empirical version. S Here he builds on distinctions introduced by B. Williams, 'Which Slopes Are Slippery?' in Moral Dilemmas in Modern Medicine, ed. M. Lockwood (Oxford: Oxford University Press, 1985): 126. 9 Walton, O.C., 127. 10 Walton, O.C., 143. I I Walton, o.c., 170. 12 Walton, (J.e., 72, quoting B.N. Waller, Critical Thinking (Englewood Cliffs, NJ: Prentice Hall, 1988): 179. 13 Walton, (J.C., 127. 14 Walton, O.c., 127-8. 15 Walton, O.C., 223. 16 Williams,o.e., 127-8. 17 In van der Burg, o.e., I have tried to lay a groundwork for such an analysis. Cf. the cri- tiques by Benjamin Freedman, Raymond J. Devettere, and David Ozar in The Journal of Clinical Ethics 3(1992)4,293-302. Devettere's critique that my analysis does not adequately address Aristotelian phronesis makes clear that on some points this groundwork should be further expanded and clarified. 18 Walton, o.C., 104. 19 Cf. Walton, (J.e., 33, 212. 20 For a historical sketch see G.K. Kimsma and E. van Leeuwen, 'Dutch Euthanasia: Back- ground, Practice, and Present Justifications' Cambridge Quarterly of Healtheare Ethics 2(1993)1, 19-35. 21 Admittedly, these two scenarios are generali- zations, but they seem to reflect the two basic outlooks. I hope to further explore this issue elsewhere. 22 The concept of a 'winding road' is introduced by 1. de Beaufort, 'Op weg naar het einde?' in Euthanasie: Knelpunten in een diseussie, ed. G.A. van der Wal (Baarn: Ambo, 1987), I. 23 Just one quotation that shows how much mis- understanding there has been of the Dutch situation. The beginning of a recent newspaper article seems to sum up the slippery slope perfectly: 'Four months after the Dutch parliament effectively decrimi- nalized euthanasia for adults, pediatri- cians in Holland are seeking similar permission for the mercy killing of se- verely handicapped newborn babies: Van der Burg's failure to deal with this most obvious counterexample to his thesis, drawn from his own backyard, is disquieting. (Benjamin Friedman, 'The Slippery Slope Reconstructed: Re- sponse to van der Burg', The Journal of Clinical Ethics 3(1992)4, 297) My response is that this is not a counterexam- ple because it is not a slippery slope, but a long and winding road. The Dutch pediatricians have been struggling with the problem of the severely handicapped newborns for at least ten years. The report referred to is the result of this long and intense discussion. Thus in a time sequence it is not, as suggested, the next step after the parliament's decision. Moreover, at the time of this newspaper ar- ticle (Vancouver Sun, 8 August 1992), the Dutch parliament had not even made such a Review of Douglas Walton 229 decision; this only passed the Lower Chamber in February 1993. Even then, it still did not formally decriminalize euthanasia: voluntary euthanasia will not be prosecuted under certain conditions, but it still remains a crime on the books. 24 I am grateful to Chris Ciecelski-Carlucci and the editors of lnfonnal Logic for their helpful comments. WIBREN VAN DER BURG CENTER FOR BIOETlflCS AND HEALTH lAW DEPARTMENT OF PHILOSOPHY UNIVERSITY OF UTRECHT HElDELBERGLAAN 8 3584 CS UTRECHT, THE NETHERLANDS :J