ISSA Proceedings 2014 ~ The Disguised Ad Baculum Fallacy Empirically Investigated – Strategic Maneuvering With Threats

Ad baculum threats can be seen as a mode of strategic maneuvering which takes on a reasonable appearance in real life situations when it mimics, legitimate pragmatic argumentation. In this paper the hypothesis was tested that ad baculum fallacies are seen as less unreasonable than clear cases when they are presented as if they are well-meant advices in which the speaker cannot be held responsible for the occurrence of the unpleasant consequences if he does not get his way.

Keywords: argumentum ad baculum, pragma-dialectics, pragmatic argumentation, strategic maneuvering

1. The argumentum ad baculum in the standard theory of pragma-dialectics
Threatening the other discussion party with negative, unpleasant consequences – for instance, by threatening him with physical violence or (more subtly) by threatening him implicitly with sanctions – if that party is not willing to refrain from advancing a particular standpoint or from casting doubt on a particular standpoint, is an outspoken example of a fallacy (“Of course, you can hold that view, but then you should realize that it will very hard for me to control my men in response to you”). Not surprisingly, this particular type of fallacy (conventionally named the argumentum ad baculum or the ‘fallacy of the stick’) has become firmly incorporated in the traditional lists of fallacies presented in introductory textbooks in (informal) logic and argumentation (cf. Walton 2000).

Seen from the perspective of the standard theory of pragma-dialectics (van Eemeren & Grootendorst 1992; 2004), the argumentum ad baculum is an example of fallacies violating the Freedom Rule (i.e. the rule for governing the first stage of a critical discussion, the confrontation stage, where standpoints are put forward by the protagonist and doubt or criticism are raised by the antagonist, in short: the stage where the difference of opinion is expressed) because, by threatening the other party and putting pressure upon him to silence and to close his mouth, the inalienable right of a discussion party to put freely forward standpoints or cast doubt on standpoints is severely hampered and restricted. As a result, a full-blown discussion hardly gets off the ground, ruling out the possibility of a resolution of the difference of opinion on the merits.

Based on the consistent results of a 13 year-lasting, comprehensive empirical research project concerning the judgments of ordinary arguers of the reasonableness of fallacious and non-fallacious discussion contributions, entitled Conceptions of Reasonableness, it can safely be concluded that ordinary arguers deem fallacious contributions as unreasonable moves, while they evaluate sound contributions as reasonable (van Eemeren, Garssen & Meuffels 2009); compared with the unreasonableness of the 24 investigated fallacies in that project (such as the ad hominem, the ad misericordiam, evading the burden of proof, the ad populum, the ad consequentiam and so on), the ad baculum fallacy – the particular fallacy we will focus on in this paper – was judged as the least reasonable discussion move (cf. van Eemeren, Grootendorst & Meuffels 1999).

From the empirical data collected in the project Conceptions of Reasonableness it can be inferred that ordinary arguers know (at least on a pre-theoretical level) where precisely to trace the boundaries of dialectical rationality; thus, at least to a certain extent, ordinary arguers are aware of their dialectical obligations. Moreover, ordinary arguers also expect that their interlocutors apply similar norms and criteria for evaluating the reasonableness of discussion contributions as they themselves do, upholding more or less the same standards of dialectical reasonableness. Last, so can be inferred from the results of our empirical research that formed a sequel of the above mentioned project, ordinary arguers use the concept of ‘reasonableness’ not only in a descriptive, but also in a normative sense: the discussant who violates one of the rules for critical discussion and thus does not observe the critical ideal of dialectical reasonableness, can be held accountable and reproached for violating commonly shared norms incorporated in the rules for critical discussion (van Eemeren, Garssen & Meuffels 2012).

2. The argumentum ad baculum in the extended theory of pragma-dialectics
All these firmly established empirical facts, however, seem at first sight not quite in line with the (supposed) frequency of the ad baculum fallacy in everyday argumentative discourse: why ever would ‘rational’ discussants use hardly efficient means like the ad baculum fallacy, a discussion move they can know and expect to be denounced by the other discussion party? Why ever would they portray themselves as being unreasonable by openly deviating from the rules of critical discussion, in the knowledge that this will make their discussion move non-persuasive in the end? Part of an answer to this paradox can be found in the so called extended standard theory of pragma-dialectics, in which a rhetorical component of effectiveness has been added to and integrated within the dialectical framework of classical, standard pragma-dialectics (van Eemeren 2010).

In their aim to be effective, discussants will maneuver strategically in such a way that they will try to achieve their dialectical goal – keeping to the rules of critical discussion – while simultaneously trying to realize their rhetorical goal: winning the discussion by having their standpoint accepted by the other party. Balancing these two objectives of dialectical resolution-oriented reasonableness and rhetorical effectiveness and trying to reconcile the simultaneous pursuit of these two aims, which may be at times at odds, the arguers make use of what can be called strategic maneuvering: a discussant tries to steer and maneuver the discussion to his advantage like a ship maneuvers for the best position in a sea battle (van Eemeren 2010: 40).

In itself there is nothing wrong with wanting to win a discussion, but trying too hard can lead to a derailment: if arguers allow their commitment to having a reasonable exchange be overruled by their eagerness for achieving effectiveness, their strategic maneuvering has been derailed. Viewed from this perspective, fallacies are derailments of strategic maneuvering that involve violations of critical discussion rules. By violating the rules for critical discussion the argumentative move they have made hinders the process of resolving a difference of opinion on the merits and so their strategic maneuvering must be condemned as fallacious.

Derailments of strategic maneuvering may easily escape attention of the interlocutors because deviations of the rules of critical discussion are often hard to detect since none of the parties in the discussion will be keen on portraying themselves as being unreasonable – if only because this will make their contribution ineffective in the end. So arguers will most likely try to stick to the established dialectical means for achieving rhetorical objectives which are possibly at odds with the dialectical rationale for a certain discussion rule, and “stretch” the use of these means so much that the fallacious maneuvering is also covered (van Eemeren 2010: 140).

As a consequence, derailments of strategic maneuvering can be very similar to sound instances of strategic maneuvering, so that in practice it is not always crystal clear where precisely the boundaries between sound and fallacious strategic maneuvering are to be found: the discrimination between fallacious and sound modes of strategic maneuvering is not a simply black or white issue. The various modes of strategic maneuvering that can be distinguished in argumentative reality can be imagined as representing a continuum ranging from evidently fallacious to evidently sound strategic maneuvering. This also goes for strategic maneuvering with particular variants of the argumentum ad baculum: at the one pool one can distinguish straightforward, clear-cut cases of illegitimate, fallacious ad baculum moves, subsequently a grey zone of argumentative threatening moves whose soundness or fallaciousness is not immediately clear, and at the other pool evidently legitimate, sound uses of threats (for instance, at the breakfast table when one authoritative party (the parents) threatens the other party (the child) with sanctions if she refuses to obey).

In the project Conceptions of Reasonableness, purposely, only clear cases of fallacies had to be judged by the participants: after all, the aim of that project was to test the conventional validity of the pragma-dialectical discussion rules (i.e. investigating whether the norms of ordinary arguers when evaluating the soundness of argumentative discourse are in agreement with the critical norms of pragma-dialectics); it was certainly not the aim of that project to investigate the factors that could influence the identification and recognition of fallacious or sound discussion moves. As said before, in everyday argumentative practice discussants maneuver strategically, attempting to hide and mask clearly unreasonable moves – like the ad baculum fallacy – by presenting these moves in such a way that they mimic and look like reasonable moves. We conjecture that one of the ways to disguise the ad baculum fallacy is to present this move as a well-meant advice backed up by legitimate pragmatic argumentation in which the speaker cannot be held responsible for the occurrence of the unpleasant consequences if he does not get his way. This hypothesis was tested systematically in two experiments.

3. Pragmatic argumentation, advising and threatening
The soundness of argumentation depends – among other things – on how it employs one of the possible argument schemes.

In pragmatic argumentation, which is a subtype of causal argumentation, the standpoint recommends a certain course of action (or discourages a certain course of action) and the argumentation consists of summing up the favorable respective unfavorable consequences of adopting that course of action (“You shouldn’t drink too much alcohol, because it leads to long-term health problems”).

The pragma-dialectical characterization of the argument scheme of pragmatic argumentation is as follows:
1. Standpoint: Action X should be carried out
1.1. Because: Action X will lead to positive result Y
(1.1’) And: (Actions of type X [such as X] that lead to positive results of type Y [such as Y] must be carried out)

Pragmatic argumentation can only succeed if the causal relation between the two elements concerned (X is the cause of Y; cf.: “too much alcohol consumption leads to health problems”) is evident and if the positive (or negative) value of the consequence Y (i.e. “having health problems is undesirable”) speaks for itself or is immediately recognized as such. In case of the ad baculum threat the other party is put under pressure by pointing or hinting at negative consequences for the other party if that party does not give in; pragmatic argumentation and the ad baculum move are thus in argumentation-theoretical respects alike in the sense that in both moves the (un)desirability of the consequences of a cause, event or act are being exploited. However, in contrast with pragmatic argumentation, the (implicit or explicit) consequences of an ad baculum move are without exception negative (in certain circumstances even frightening and fear-inducing).

Pragmatic argumentation is by convention associated with the speech act of advising (or warning) (cf. van Poppel 2013): in order to make an advice or warning acceptable for an audience (“You should do…” or “You shouldn’t do …”), pragmatic argumentation is characteristically adduced. Both the act of advising and the act of threatening are speech acts that can be classified – looking at their (primary) illocutionary goal – as directives; moreover, both speech acts have felicity conditions in common (such as the preparatory condition concerning the authoritative status of the source of the advice/threat).

Mimicking the ad baculum as a well-meant advice that is in the interest of the hearer would certainly not be sufficient – as we conjecture – for the persuasiveness of such disguised form of threat. Despite all the similarities and resemblances between the uses of the pragmatic argument scheme adduced in advises and threats, there is one crucial difference between these two speech acts: in case of an advice or warning (“You shouldn’t drive so speedy, darling. It’s raining!”) the other party in the discussion has full freedom and responsibility for the occurrence (or non-occurrence) of the effect Y (in the causal relation: If X, then Y); however, in case of a threat the party who advances the threat can be completely held responsible for the occurrence of the negative outcome (“If you still persist in that awful behavior, I have to dismiss you”). The secondary illocutionary goal of a threat can thus be conceived as that of a commissive (i.e. a commitment undertaken by the speaker vis-à-vis the listener to do something and act according to what is explicitly said or implied by what is said).

In order to disguise the ad baculum in a strategically effective way and to make this fallacious move look like a legitimate discussion move, it is vital for the speaker to suggest/hint that evidently not he or she, but another party or event outside the discussion can be held accountable for the occurrence of the undesirable, negative outcome. Expressed differently: the causal relation in the pragmatic argument scheme (X is the cause of the effect Y, or: the act of X is leading to the consequence Y) is deceitfully represented and treacherously exploited in such a way that the arguer (the person who advances the threat) cannot be held accountable for the occurrence of the negative effect Y: it is after all not the arguer but a party outside the current discussion that can be blamed.

To illustrate these points, take the following two examples (the first is an example of an openly, straightforward clear-cut case of an argumentum ad baculum, the second an example of a disguised ad baculum – disguised according to the conjectural ideas above). Suppose two neighbors (Sally and John) argue about the annoying barking of John’s dog. Sally is completely fed up with that barking, especially in the night.

Sally: You should learn that dog not to bark at night; every night I wake up because of that terrible noise.
John: What nonsense, he really doesn’t bark that much at all.
Sally: If you keep saying that, I’ll harm him.

Sally’s last move is forthright ad baculum: she explicitly commits herself to ‘kill’ the dog if John refuses to take any measures. But Sally could have chosen to present her last move in a strategically, perhaps more effective way – more effective as we predict –, namely as a well-meant advice, disguising the threat but without undoing it:

Sally: You should learn that dog not to bark at night; every night I wake up because of that terrible noise.
John: What nonsense, he really doesn’t bark that much at all.
Sally: I would strongly advise you to take effective measures to stop that awful barking. You wouldn’t like it if somebody would harm your beloved dog, wouldn’t you?

In the two experiments reported in this paper, the crucial contrast is that between the (perceived) unreasonableness of straightforward ad baculum moves and the unreasonableness of disguised ad baculum moves. In all cases we present instantiations of the disguised fallacy as a well-meant advice that is in the interest of the addressee, making use of indicators of the speech act of ‘advising’ such as: “I would advise you…”; “It would be wise if you….”; “If I were you, I would…”; “If I were in your position, I would…”; “I would recommend you …”, If you are asking me, I would I think ….”

The arguer, however, has still various other – perhaps strategically effective – presentational devices at his disposal to mask other aspects of the threat, for instance devices to undo the inherent, annoying pressure of the ad baculum move, which is at odds with someone’s personal freedom. To guarantee that it is absolutely not his intention to threaten the opponent and to put pressure on him, the arguer can strategically emphasize that the other party is “totally free to decide whatever she wants”: “Of course you are absolutely free to decide whatever you want, but if I were in your position …”; “It’s totally up to you, but I would advise you…”. In the two experiments we conducted, we presented (hypothetical) discussion fragments to the participants in which – in case of the disguised ad baculum – only indicators of the speech act of ‘advising’ were used.

4. The experiment
In the current study we tested the following main hypothesis:

Ad baculum fallacies are judged as less unreasonable than clear-cut, straightforward cases of ad baculum moves when they are presented as if they are well-meant advices in which the speaker can’t be held responsible for the occurrence of the unpleasant consequences if he does not get his way.

The experimental (Dutch) subjects (Ss) were exposed to 42 discussion fragments; some contained fallacious moves, others did not. In each dialogue, the Ss had to rate the (un)reasonableness of the last contribution to the discussion on a 7-point Likert type of scale, ranging from ‘very unreasonable’ ( = 1) to ‘very reasonable’ ( = 7).

4.1 Material
42 discussion fragments were constructed, in which 7 different types of fallacious and non-fallacious discussion contributions occurred; each type was represented by 6 instantiations:

(1) straightforward, clear-cut cases of ad baculum moves,
(2) disguised ad baculum moves,
(3) sound, i.e. reasonable moves (not based on a pragmatic argument scheme),
(4) sound, i.e. reasonable moves (based on a pragmatic argumentation scheme),
(5) the circumstantial variant of the ad hominem fallacy,
(6) the tu quoque variant of the ad hominem fallacy,
(7) the abusive variant of the ad hominem fallacy.

The general structure of these discussion fragments was fixed: all fragments consisted of 3 turn dialogues between two discussants; each fragment was preceded by a short contextual description to ensure that the Ss interpreted the fragment in a more or less homogeneous way. Just like in our previous studies we did not include loaded topics; we tried to keep the dialogues as simple as we could and avoided humorous situations or elements that could otherwise distract our respondents.

In the first turn, the protagonist put forward a standpoint, supported by an argument. In the second turn, the antagonist made explicitly or implicitly clear not to accept that standpoint, backed up by an argument. In the last turn (in case of a straightforward ad baculum), the protagonist implicitly and indirectly threatened the other party by pointing at negative consequences if he does not get his way, like this:

(1) Straightforward ad baculum
Employer and employee during a performance interview

Employee: I think it is time for a promotion. My work really improved much and I receive a lot compliments from my colleagues.
Employer: I don’t agree, there are a lot of points for improvement.
Employee: Well, you may maintain that point of view, but I know about your creative way of making your tax returns and you do not want that out in the open.

Notice that in the example above, as in all the other 5 instantiations of the straightforward ad baculum move, the protagonist threatens the other party implicitly with non-physical consequences that are indirectly put forward, i.e. not explicitly spelled out. Making use of such indirect, non-physical consequences in spelling out the negative consequences makes it much harder for us to confirm our main hypothesis, compared with physical, direct ad baculum moves.[i] The following is an example of a disguised ad baculum, constructed according to the theoretical insights outlined above:

(2) Disguised ad baculum
The stage-manager and the key actress are discussing the suitability of her costume.

Stage-manager: This costume is really splendid, it does perfectly fit with the role.
Actress: I hate it!! That dress makes me look awfully fat!
Stage manager: I would advise you just to put it on, it’s really a nuisance if another main actress has to be looked for.

Once again, in the current experiment the disguised ad baculum is always presented (in the 6 instantiations) as an explicit advice which is in the interests of the addressee, accompanied by an explicit indicator of the speech act ‘advising’.

For the purpose of constructing a base line for comparisons and contrasts between fallacious and non-fallacious moves, in 6 dialogues ‘normal’ non-fallacious reactions were included (reactions, however, in which no pragmatic argumentation was used, but other argument schemes). For an example of this type of dialogue, see (3):

(3) Sound argumentation (in which the pragmatic argument scheme is not used)
A young couple discusses their opinions after seeing the stage play.

Alissa: What a wonderful play! The actors had a very professional mimic and attitude.
Mark: I didn’t like the play at all, the topic was very boring.
Alissa: No, on the contrary, that topic wasn’t boring at all! It covered all the facets of real life and it was highly instructive.

In (4), an example of sound argumentation in which pragmatic argumentation is used, is presented. Evidently, such examples are relevant for an appropriate contrast between the (perceived) (un) reasonableness of the fallacious use of pragmatic argumentation (as is the case in disguised ad baculum moves) and the (perceived) (un) reasonableness of sound, non-fallacious use of pragmatic argumentation:

(4) Sound argumentation (in which the pragmatic argument scheme is used)
Pim and Anke in their car on the highway, discussing the speed limits:

Anke: Please slow down! The upper limit here is 100 km.
Pim: Don’t be so nervous, everybody is driving faster so it doesn’t really matter.
Anke: If I were you, I would keep up to the maximum speed; soon you will be caught and get a ticket.

Three types of filler items were included as well: 6 dialogues containing a tu quoque fallacy, 6 dialogues containing a circumstantial ad hominem fallacy, and 6 dialogues containing an abusive ad hominem fallacy (for concrete examples, see van Eemeren, Garssen & Meuffels 2009). These fillers acted as ‘gate keepers’: we included these kinds of fallacies in the questionnaire because, given the consistent results reported in the Conceptions of Reasonableness project, we know exactly what to expect when it comes to reasonableness judgments about these fallacies (namely, the abusive attack is judged as a very unreasonable move, whereas the circumstantial as well as the you too-variants tend to be judged as reasonable moves).[ii] If these expectations would not be met in the current study, this would imply a serious threat to the validity of the present investigation. A second reason for including these fillers was to mask the precise aim of our research focusing on ad baculum fallacies. Varying the type of fallacy made it more difficult for our respondents to infer a pattern in the material and to guess what our experiment was aimed at.

4.2 Participants
A total of 93 secondary school students (pre-university level, ranging in age from 14 to 18; M = 15.94; SD = .75; 41% male, 59% female) took part in the pencil-and-paper test during regular class hours. Some of them knew the term fallacy, but none of them had received any systematic education regarding argumentation.

4.3 Statistical analysis
The data were analyzed by means of a multivariate analysis of variance (‘mixed model’ approach for repeated measurements), with ‘subject’ and ‘instantiation’ as random factors and the variable ‘type of fallacy’ as a fixed factor. The random factor ‘instantiation’ is nested within the levels of the fixed factor ‘type of fallacy’, whereas the random factor ‘subject’ is fully crossed with the random factor ‘instantiation’ and the fixed factor ‘type of fallacy.’ The statistical consequence of this design is that – instead of ordinary F-ratio’s – so-called quasi F-ratios have to be computed (denoted as F’), while the degrees of freedom have to be approximated (see Clark 1973).

4.4 Results
Looking first at the fillers (Table 1), it is evident that the present results are in line with the results we found in our previous studies conducted in the project Conceptions of Reasonableness. The abusive fallacy is again judged to be most unreasonable, next the circumstantial attack and last the tu quoque fallacy, both of which tend again to be viewed as reasonable moves. Moreover, the perceived unreasonableness of the straightforward ad baculum fallacy as well as the judged reasonableness of sound argumentation is equally well in accordance with the empirical findings in of Conceptions of Reasonableness. In sum, the reasonableness scores depicted in Table 1 are a positive indication for the validity of the data.

Do the respondents – as predicted in our hypothesis – regard ‘hidden’ ad baculum moves which mirror well-meant advices supported by pragmatic argumentation indeed less unreasonable as straightforward, clear-cut cases of ad baculum? In Table 2 the relevant data are presented.

The average reasonableness scores pertaining to the four types of moves in Table 2 proved to differ statistically from each other: F’ (3,22) = 14.27, p < .01; η2 = .36. By means of three orthogonal post hoc comparisons we contrasted, first, sound non-pragmatic argumentation with sound (pragmatic) argumentation, but no statistical difference could be found (F‘ (1,22) = 1.43, n.s.). Second: the disguised ad baculum differed significantly from the average of the sound non-pragmatic argumentation and the sound pragmatic argumentation: F’ (1,22) = 6.64, p < .07; η2 = .03. Last, and most important for our hypothesis, the disguised ad baculum was indeed found to be less unreasonable than the straightforward ad baculum: F‘ (1,22) = 10.97, p < .01, η2 = .10. The difference between these two fallacious threats (1.58) is considerable, given the range of a 7-point scale. Our respondents clearly judged the straightforward ad baculum threat as an unreasonable argumentative move, but when it comes to judging the disguised form of this fallacy they are clearly in doubt: overall this fallacious move is judged as neither unreasonable nor reasonable.

5. Replication
In order to be able to generalize the results with more confidence, a replication was carried out, making use of different messages and different subjects. 128 students (high vocational education; age range 17-31 (M = 20.59; SD = 2.66)) were exposed to 42 different, but equivalent messages as in the experimental study above. Instead of the circumstantial variant of the ad hominem, we now used the fallacy of shifting the burden of proof and – once again – the tu quoque variant of the ad hominem as “gatekeepers” of the validity. This time each type of fallacy and sound argumentation was represented not by 6, but by 7 instantiations.

The average reasonableness scores for the gatekeepers were again in line with the expectations, derived from the consistent results in the Conceptions of Reasonableness project (clear case ad baculum: M = 2.74; SD = 0.65; shifting the burden of proof: M = 3.06; SD = 1.05; tu quoque: M = 4.12; SD = 0.84; sound (non-pragmatic) argumentation: M = 5.59; SD = 0.59). The statistical results of the replication are also in accordance with those of the original experiment. Once again, there were statistically significant differences between the four types of reactions depicted in Table 3: F’(3, 25) = 16.65, p < .001, η2 = .40.

The orthogonal post hoc contrast between sound non-pragmatic argumentation and sound pragmatic argumentation was once again found to be statistically not significant: F’(1,25) = 0.00, p = .99. Just as in the previous experiment, the disguised ad baculum fallacy differed significantly from the average of the two types of reasonable argumentation: F‘(1,25) = 18,49, p < .001. Last, the disguised ad baculum was once again found to be substantially less unreasonable than the explicit variant of the ad baculum fallacy: F‘(1,25) = 4.33, p < .05.

6. Conclusion
The empirical results of the original experiment and those of the replication are quite similar and in line with our theoretical expectations: Ordinary arguers clearly reject straightforward ad baculum moves; disguised forms of such moves are judged substantially less unreasonable by our experimental subjects, since these moves take on a reasonable (but treacherous) appearance – indeed, the Latin word fallax means deceptive or deceitful – when they are presented as if they are well-meant advices backed up by pragmatic argumentation in which the speaker cannot be held responsible for the occurrence of the unpleasant consequences if he does not get his way.

In earlier empirical studies in which we investigated strategic maneuvering with abusive ad hominem attacks, we showed that direct attacks are judged as less unreasonable when they are presented as if they are critical questions pertaining to the argument scheme for authority argumentation (van Eemeren, Garssen & Meuffels 2010); we coined that strategic effect the mimetic effect. Given the current empirical findings concerning ad baculum fallacies it can be concluded that this mimetic effect is not specifically bound to strategic maneuvering with ad hominem fallacies, but can be generalized to other types of fallacies.

Another remarkable empirical finding that is strikingly similar in both studies is the size of this mimetic effect: the disguised forms of both fallacies (i.e. the ad hominem as well as the ad baculum) are evidently not judged as fully or fairly reasonable moves; the judgments center around the neutral midpoint of 4 on the 7-point scale. So, ordinary arguers are clearly in doubt and are quite uncertain when it comes to judging the reasonableness of these disguised forms. The appearance of a certain, modest degree of reasonableness is presumably sufficient for arguers to get away with such treacherous moves in argumentative discussions.

NOTES
i. In the experiments pertaining to the unreasonableness of different forms of ad baculum fallacies (such as threatening with physical consequences vs. threatening with non-physical consequences; and threatening in a direct way vs. threatening in an indirect way) it was found that threatening with physical consequences was judged most strictly, while indirect threatening was deemed to be the least unreasonable move (see van Eemeren, Grootendorst & Meuffels 1999). So, by making use of only indirect forms of straightforward ad baculum fallacies in the present experiment, a far too easy confirmation of our hypothesis is avoided.
ii. That the circumstantial as well as the you too variants tend to be judged as reasonable moves is only the case when participants have to judge the reasonableness of these fallacies presented in unspecified contexts. When these two types of fallacies are presented in a scientific context, these variants of ad hominem are deemed to be unreasonable, like the abusive variant.

References
Clark, H. H. (1973). The language-as-fixed-effect fallacy: A critique of language statistics in psycholinguistics. Journal of verbal learning and verbal behavior, 12, 335-359.
Eemeren, F. H. van (2010). Strategic maneuvering in argumentative discourse. Extending the pragma-dialectical theory of argumentation. Amsterdam-Philadelphia: John Benjamins.
Eemeren, F. H. van, Garssen, B., & Meuffels, B. (2009). Fallacies and judgments of reasonableness: Empirical research concerning the pragma-dialectical discussion rules. Dordrecht: Springer.
Eemeren, F. H. van, Garssen, B., & Meuffels, B. (2012). The disguised abusive ad hominem empirically investigated; Strategic manoeuvering with direct personal attacks. Thinking and Reasoning, 18(2012), 2, 344-364.
Eemeren, F.H. van, Garssen, B., & Meuffels, B. (2012). The Extended Pragma-Dialectical Argumentation Theory Empirically Interpreted. In: Frans H. van Eemeren & Bart Garssen (eds.), Topical Themes in Argumentation Theory: Twenty Exploratory Studies (Argumentation Library, vol.22). Dordrecht: Springer, 323-343.
Eemeren, F. H. van, & Grootendorst, R. (1992). Argumentation, communication, and fallacies: A pragma-dialectical perspective. Hillsdale, N.J.: Erlbaum.
Eemeren, F. H. van, & Grootendorst, R. (2004). A systematic theory of argumentation: The pragma-dialectical approach. Cambridge: Cambridge University Press.
Eemeren, F.H. van, Grootendorst, R., & Meuffels, B. (1999). De onredelijkheid van de ad baculum-drogreden. Taalbeheersing, 21, 1, 29-48.
Walton, D. N. (2000). Scare tactics: Arguments that appeal to fear and threats. (Argumentation Library, Vol. 3). Dordrecht: Springer.
Poppel, L. van (2013). Getting the vaccine now will protect you in the future! A pragma-dialectical analysis of strategic maneuvering with pragmatic argumentation in health brochures. Diss. University of Amsterdam.