Handbook of the History of Logic: - Fordham University Faculty

564 Simo Knuuttila
means, consenting to the antecedent (goal) implies consenting to the consequent
(a means). 232
Some twelfth-century authors realized that even though willing an end effectively
implies that a necessary means to it is also willed, a rule (W), analogical
to (5) cannot be applied without restriction to efficient will in the form that if
Socrates wills that p and he knows that there is a valid consequence from p to q,
he also wills that q. The principle holds about the end-means relationship, but the
necessary means are not the only consequents of what is willed. Peter of Poitiers
formulated a counter-example as follows: if S repents of a sin, S is guilty of a sin,
and S wills to repent of a sin, but S does not will to be guilty of a sin. 233 Stephen
Langton’s counter-example was of the same type: if a man visits his sick father,
the father is sick. This man wills to visit his sick father, but it does not follow
that he wills the father to be sick. 234
When it was realized that one cannot apply (W) to effective will without qualification,
medieval thinkers took an interest in finding cases in which a rational
agent does not will the consequent of what he or she wills. The twelfth-century
examples just mentioned specify the case in which to will something in certain circumstances
implies that those circumstances prevail, although this is something
which the agent does not will. This exception to (W) was thought philosophically
interesting and was also used in later medieval deontic logic. Contrary to
what Abelard thought, it was realized that (4) and (5) cannot be applied without
qualification to the notions of obligation and permission. The reasons were similar
to those which demanded the qualification of (W). Another relevant but more
problematic idea was that a separately unwanted side-effect of what is willed is
willed more indirectly than a separately unwanted means to an end. This line of
thought forms the background to Thomas Aquinas’s distinctions between directly
and indirectly willed things. 235
One can follow the discussion of the question of how (W) should be qualified
with respect to will by tracing the comments on the example of a man who ‘wills to
be in the mud and have 100 marks’ (in luto esse cum 100 marchis). This example
was considered ambiguous, and it was possible to interpret it (i) as a case of willing
reluctantly to become dirty as a means of receiving 100 marks, (ii) as a case of
willing to become dirty as a side-effect or (iii) willing something in a situation in
which one would become dirty in any case. The third alternative was particularly
relevant to later discussions of conditional obligations.
The most extensive fourteenth-century discussion of the logical properties of
deontic concepts is to be found in the first question of Roger Roseth’s Lecture on
232Ethics 8.21-10.2; 14.14-19; 16.11-18.
233Sententiae IV.16 in Patrologia latina, ed. J.P. Migne, 211, 1199.
234Quaestiones theologiae, quoted in R. Quinto, ‘Die Quaestiones des Stephan Langton über
die Gottesfurcht’, Cahiers de l’Institut du Moyen-Âge Grec et Latin, Université de Copenhague
62 (1992), 77-165, at 129-130.
235See, e.g., Summa theologiae II.2.64.7-8; De malo 2.1. See also Knuuttila 2006, 205-18.