and memory in ‘‘Late Scholastic Memoria.’’ See the articles on episodic
and semantic memory in Dudai, Memory from A to Z: these clarify impor-
tantly how modern analysis of memory has changed so radically from the
ancient that it is difficult (though necessary) for historians even to use the
same lexicon. See also Dudai and Carruthers, ‘‘The Janus Face of
Mnemosyne.’
39. This crucial, much-cited passage, is also found in Aristotle, Posterior Analytics,
I I , xix (100a).
40. As is apparent from Tracy, the notion of ‘‘embodiment’’ is elementary in
Aristotle, and Aquinas rejects any modification of this idea in the area of how
human beings know, specifically the implications in Avicenna and Averroe¨s
that direct knowledge of abstractions is possible. See Kenny, ‘‘Intellect and
Imagination in Aquinas,’’ and Aquinas on Mind, 89–99.
41. Thomas Aquinas, In De anima commentarium, Lectio 13 on Book I I I , par. 792.
I quote the translation by Kenelm Foster; the Latin text is that of the Marietti
edition (Turin, 1949).
390
Notes to pp. 69–71
42. Sorabji, Aristotle on Memory, 14–17; the image, from Aristotle’s De anima, is
commented on by Thomas Aquinas in par. 26 of Lectio 2 on Book I.
43. Thomas Aquinas, In De anima, Lectio 14 on Book I I , par. 417.
44. Thomas Aquinas, In De anima, Lectio 1 on Book I I I , par. 570–574.
45. Thomas Aquinas, In De anima, Lectio 14 on Book I I , par. 418.
46. Thomas Aquinas, In De anima, Lectio 14 on Book I I , par. 417.
47. Thomas Aquinas, In De anima, Lectio 14 on Book I I , par. 416.
48. Thomas Aquinas, In De anima, Lectio 14 on Book I I , par. 406–414.
49. Thomas Aquinas, In De anima, Lectio 24 on Book I I , par. 553–4.
50. Thomas Aquinas, In De anima, Lectio 24 on Book I I , par. 552. The nature of
Aristotelian ‘‘hylemorphism,’’ that is, the idea I have been describing that
perception occurs by causing physical changes in the organ that make it ‘‘like’’
what it perceives, is analyzed by Tracy, who sees it as a consequence of
Aristotle’s idea that the body is a balanced compound of the four elements,
animated by a soul. These elements and their associated qualities (heat, cold,
moisture, dryness) occur in paired opposites, which react when contact
between them is established. ‘‘If the objective quality is strong enough, the
qualitative change which it sets up in the medium evokes a corresponding
change in the sense organ, i. e. the organ responds in the direction of that
quality in proportion to its intensity. In so doing, the organ becomes like the
objective quality’’ (Tracy, 207). But Burnyeat disagrees strongly with this
interpretation, and argues that for Aristotle ‘‘spiritual’ change does not in fact
entail material change in the organ. See his analysis of Aquinas’s commentary
(arguing that Aquinas understood Aristotle well on this point), ‘‘Aquinas on
‘Spiritual Change’ in Perception.’’ See also Nussbaum’s discussion of hyle-
morphism and Aristotle’s concept of pneuma, 146–164. Thomas Aquinas, like
Galen, located the sensus communis in the first ventricle of the brain. He also
seems to make a sharper distinction than Aristotle does, between a sense like
touch, which is directly affected by its object, and sight, whose ‘‘likeness’’ is
‘‘spiritual,’ in the way I am attempting to describe. For Thomas, sight is ‘‘like’’
its object as the imprint of a seal in wax is ‘‘like’’ the sealing-ring, an image
which links the product of sight directly to the nature of the memorial
phantasm. Clearly, the analogies between the external and inner ‘‘eyes’’ were
far-reaching in Thomas’s thought. See also Kenny, Aquinas on Mind, 88–117,
on sensations, phantasms, and human knowledge of universals and
particulars.
51. Thomas Aquinas, In De anima, Lectio 6 on Book I I I , par. 669.
52. Thomas Aquinas, In De anima, Lectio 13 on Book I I I , par. 792.
53. Quoted from Averroe
¨s, Epitome of the Parva naturalia, II, i; trans. Blumberg, 30.
54. For analysis of Aristotle’s account, see Sorabji, Aristotle on Memory, 14–17, and
Burnyeat, ‘‘Aquinas on ‘Spiritual Change.’’’ An English translation of
Aquinas’s commentary on Aristotle’s De memoria is in Carruthers and
Ziolkowski, eds., The Medieval Craft of Memory, 153–188; see also the trans-
lation of Albertus Magnus’s commentary on the same text, The Medieval Craft
of Memory, 118–152.
Notes to pp. 71–76
391
55. Thomas Aquinas, In De anima, Lectio 2 on Book I , pars. 19–20; see also ST Ia,
Q.76, art. 1.
56. Thomas Aquinas, In De anima, Lectio 12 on Book I I I , par. 783.
57. Thomas Aquinas, In De anima, Lectio 12 on Book I I I , par. 782.
58. Sorabji, Aristotle on Memory, 17.
59. Thomas Aquinas, In De anima, Lectio 12 on Book I I I , par. 776.
60. Thomas Aquinas, In De anima, Lectio 12 on Book I I I , par. 773.
61. ST Ia, Q. 85, art. 1, obj. 4.
62. ST Ia, Q. 75, art. 5; also Q. 76 passim, but esp. arts. 1 and 3.
63. See Paradiso, 29, 76–81.
64. Gardner, ‘‘Imagination and Memory,’’ esp. 280–282. The quotation is from 282.
This passage has received some comment from Mazzotta, who sees the ‘‘failure’’ of
Dante’s memory as part of a conscious deconstructionist enterprise in the poem
(Dante: Poet of the Desert, 260–267). Insofar as Dante’s term ‘ failure’’ is under-
stood by modern interpreters to carry a negative moral connotation, it seems to
me a wrong emphasis, and somewhat perverse, to label as a moral fault what is the
intrinsic limitation of a biological mechanism (memory). See the comments of
Freccero on ‘‘The Final Image’’ in his Dante: the Poetics of Conversion, 245–257.
65. ST Ia, Q. 77, art. 8.
66. ST Ia, Q. 79, art. 6, resp. obj. 2.
67. Sorabji discusses the difference very clearly, Aristotle on Memory, 11–12.
Aristotle’s discussions of the matter in both ‘‘On Divination in Dreams’’
and ‘‘On Dreams’’ are relevant.
68. Aristotle calls them ‘‘the movements of the senses when one is asleep’’; On
Dreams, 462a, 15–30. Averroe¨s says that sleep is the ‘‘sinking of [the] common,
perceptive faculty into the interior of the body’’; Epitome, II, ii (trans.
Blumberg, 33).
69. Avicenna, Liber de anima, I V , ii: ‘‘haec est propria prophetia virtutis imagi-
nativae’’ (Avi. Lat., vol. 2, 19, line 61).
70. Avicenna, Liber de anima, I V , ii (Avi. Lat., vol. 2, 19–20, lines 68–70):
‘‘aliquando est ab intellectibus et aliquando est a divinationibus et aliquando
est ex versibus, et fit hoc secundum aptitudinem et usum et mores.’
71. Avicenna, Liber de anima, I V , ii (vol. 2, 20, lines 71–75): ‘‘quae adiuvant
animam plerumque incognitae et plerumque sunt sicuti apparitiones subitae,
quae non sunt residentes ita ut rememorari queant nisi eis succurrerit anima
cum retentione appetita quia, quod potius agit anima, hoc est scilicet retinere
imaginationem circa genus dissimile ab eo in quo erat.’’
72. Quotations in this paragraph are all from Averroe
¨s, Epitome, II, i, p. 28. These
comments also occur in Albertus Magnus’s commentary on Aristotle’s De
memoria, evidence of how widely spread this tradition was in teaching.
Albertus follows his discussion of the image-forming ability’s role in reminis-
cence by adducing the advice for making vivid memory-images found in the
Rhetorica ad Herennium.
73. Sacks, The Man Who Mistook His Wife for a Hat, discusses the phenomenon
now restrictively called ‘‘reminiscence’’; he suggests that to restrict the term as
392
Notes to pp. 76–80
modern neuropsychology has may be an error of an overly mechanistic
science, and comments that memories may be more like scripts or scores
than algorithms; see 138–142, and the interesting case-study that is the title-
essay of the book. Certainly both Aristotelean analysis and practical mne-
monics recognized that recollection proceeds via a network of associations, the
‘‘richer’’ the better and the more secure.
74. De memoria, 449a 9ff.; Sorabji, Aristotle on Memory, 47–48.
75. Averroe
¨s, Epitome, II, i, p. 30. Aristotle stresses the emotional accompani-
ment of recollections in De memoria, 453a. This concept is close to what
modern neuropsychologists recognize as episodic memory, which they
sharply distinguish from semantic memory, a distinction not recognized in
pre-modern psychology. Ancient and medieval students were taught to
embed semantic details in episodic recollections as a mnemonic strength-
ener: see below.
76. Petrarch, Rerum memorandarum libri, I I (‘‘De memoria’’), 13: ‘‘Vidisse semel
vel audisse sat est, nunquam obliviscitur, nec res modo meminit, sed verba
tempusque et locum ubi quid primum accepit. Sepe totos dies aut longas
noctes colloquendo transegimus: audiendi namque cupidior nemo est; post
annos vero suborta earundem rerum mentione, siquid forte plus minusve aut
aliter dixissem, submissa voce confestim admonebat hoc me aut illud verbum
immutare; mirantique et unde hoc nosset percunctanti non solum tempus
quo id ex me audivisset, sed sub cuius ilicis umbraculo, ad cuius undam
fluminis, in cuius maris litore, cuius montis in vertice – longinquas enim
secum oras circuivi – me singula recognoscente memorabat.’’
77. Inst. orat., X I . ii. 20: ‘‘Ita, quamlibet multa sint, quorum meminisse oporteat,
fiunt singula conexa quodam choro, nec errant coniungentes prioribus con-
sequentia solo ediscendi labore.’’ This is a corrupted passage, though its
general sense is clear: see Russell’s note in his translation for the Loeb series,
and the variant readings recorded in Winterbottom’s Oxford Classical Texts
edition. I have quoted Butler’s old Loeb translation in this one instance;
otherwise I have cited Russell’s throughout.
78. Inst. orat., V . x. 20–22.
79. Aristotle, De memoria 451b 18ff.; in this one instance I have used the Loeb
translation by W. S. Hett.
80. Aristotle, De memoria, 451b 29; see Sorabji’s discussion of how recollecting
(the mnemonic search) differs from remembering (the object of memory),
Aristotle on Memory, 41–48.
81. Sorabji, Aristotle on Memory, 31.
82. Sorabji, Aristotle on Memory, 29–30. I discuss this important passage in the
Topics at greater length in connection with the ancient method of using
numbers as a recollective scheme; see Chapter 3.
83. ST Ia, Q. 78, a. 4, resp.
84. Quintilian comments that ‘‘we must, of course, invent’’ for ourselves the
mnemonic images we use, unlike our background places, which may be real
or fictive; Inst. orat., XI. ii. 22.
Notes to pp. 81–85
393
85. Thomas Aquinas, commenting on this passage, defines the associative nature
of recollection clearly: In De memoria et reminiscentia commentarium (here-
after cited as In de memoria comm.), Lectio VI, 379–383.
86. ST Ia, Q. 93, art. 7, ad 3. See also Q79, art. 7, ad 1.
87. ST I I –I , Q. 50, art. 5, resp.
88. See especially Rand, Cicero in the Courtroom, 25–33, 103–112; in the matter of
the cardinal virtues, Rand observes, Cicero was Thomas Aquinas’s ‘‘constant
companion and one of his respected authorities’’ (32).
89. Douay translation of Vulgate ‘‘sobrietatem enim et prudentia docet, et
justitiam et virtutem, quibus utilis nihil est in vita hominibus.’’
90. Rand, Courtroom, 26–28; see also Loeb note on De inventione. The word
phron¯esis, sometimes also translated as prudentia, has a complex medieval
history. Aristotle often links it with doxa or ‘‘judgment,’’ but says that while
animals also have phron¯esis, only humans have doxa. Sorabji translates
the word as ‘‘intelligence’’ (77–78), and the word clearly describes a sub-
conceptual and sub-judgmental ability rather like what we mean when we say
that our dog is ‘‘intelligent.’’ But in the Middle Ages, intelligentia appears as
one of the parts of prudence (along with memory and providence), following
the definition of prudence given in De inventione, II, 160. This use of
intelligentia is rendered as phron¯esis by Albertus Magnus.
91. Ad Her. I I I , 2, 3.
92. Cicero, De inventione, I I , 53, 160.
93. ST I –I I , Q. 57, art. 3, ad 3.
94. ST I –I I , Q. 57, art. 1, concl.
95. ST I –I I , Q. 57, art. 4, resp.
96. ST I –I I , Q. 57, art. 5, resp.
97. ST I I –I I , Q. 48, resp.
98. Brunetto Latini, Livres dou treśor, I I , 59, lines 2–5 (ed. Carmody, 233–234).
On prudence as containing all teaching, see II, 56–58; for example, ‘‘Mais ki
bien consire la verite´, il trovera que prudence est le fondement des unes et des
autres; car sans sens et sans sapience ne poroit nus bien vivre, ne a Dieu ne au
monde’’; II, 56, lines 4–5 (230–231). The ‘‘Seneque’’ referred to is probably
Martin of Braga (sixth century), Pseudo-Seneca, whose brief moral treatise,
‘‘Formula honestae vitae,’’ was widely known during the Middle Ages:
‘‘Si prudens est animus tuus, tribus temporibus dispensetur. Praesentia
ordina, futura praevide, praeterita recordare’’ (PL 72, 24C). I am indebted
to Paul Gehl for this citation.
99. ST I I –I I , Q. 49, art. 1. Notice how Thomas Aquinas has brought together the
texts on ethos/¯ethos from the Ethics and Metaphysics.
100. See Tredennick’s translation of the Greek text in the Loeb edition.
101. My quotation is from Rand, Cicero in the Courtroom, 61–62. A list of Thomas
Aquinas’s pagan sources is on 17–18 of this same book.
102. ST I I –I I , Q. 49, art. 1, obj. 2.
103. Tracy, Physiological Theory and the Doctrine of the Mean, 251. The influence of
Aristotle’s ethical theory, embedded as it is in emotional life as well as reason,
394
Notes to pp. 85–88
upon later antique and early Christian ethical philosophy is discussed by
Sorabji, Emotion and Peace of Mind. See particularly the discussion of
Seneca in this study, and in Nussbaum, The Ther
apy of Desire.
104. See the first sentence of Moerbeke’s text, in Thomas Aquinas, In De memoria
comm.
105. Aristotle, De memoria, 450a 25–30.
106. Aristotle, De memoria, 449b 24; for discussion see Sorabji, Aristotle on
Memory, 1–2, for other uses of hexis and pathos in this sense in De memoria.
107. Sorabji, Aristotle on Memory, 1.
108. Aristotle, Nicomachean Ethics, 1103a 17ff.
109. Aristotle, Nicomachean Ethics, 1104a 27–30. In his commentary on this
passage (In decem libros ethicorum commentarium, Bk. II, Lectio I, 247),
Thomas Aquinas notes the pun in Greek ethos, and concludes, ‘‘Just so for
us the word moral signifies something formed from habit, whenever it
pertains to vice or virtue’’ (sicut etiam apud nos nomen moralis significat
quandoque consuetudinem, quandoque autem id quod pertinet ad vitium
vel virtutem). Consuetudo is defined as synonymous with mos, ‘‘custom.’’
Furthermore, virtus moralis is described as ‘‘in parte appetitiva,’’ and as such
is based in something more like what we would call ‘‘physiological’’ than
what we might call ‘‘rational’’ or even ‘‘mental.’’
110. Tracy, Physiological Theory and the Doctrine of the Mean, 229–231; the
quotation is from 230. Tracy’s analysis has been criticized as overly materialist
by Hutchinson, ‘‘Doctrines of the Mean,’’ 19, note 4, who notes that Galen
made the same error in interpreting doctrines of Plato and Aristotle.
111. Aristotle, Posterior Analytics, I I , 19; 100a 1–5.
112. Quintilian discusses hexis as firma facilitas in Inst. orat., X , esp. parts i. 1 and
i. 59, but all of X. i is pertinent to the matter of attaining mastery of oratory,
as are his remarks on imitation in Book X. ii.
113. Norman, Learning and Memory, 74–80.
114. Thomas Aquinas, In de memoria comm., Lectio I , 298.
115. Tracy, Physiological Theory and the Doctrine of the Mean, 229.
116. Averroe
¨s II, 1, p. 30; cf. also Aristotle, De memoria, 453a 15–30.
117. ST I –I I , Q. 56, art. 5, obj. 3.
118. The Rhetorica ad Herennium and the De inventione were copied together
during the Middle Ages; the Ad Herennium was not proven to be by someone
other than Cicero until the sixteenth century, but there were suspicions
earlier. On the dissemination of these texts and their traditions during the
The Book of Memory Page 67
