Self-knowledge in Cicero and Augustine (De Trinitate 10)

1. See Porphyry, sent., 43. Augustine uses this contrast between the eye and themind several times in De trinitate; see especially trin., IX, 3, 3 (cf. X, 7, 10-8, 11, dis-cussed below) and XIV, 6, 8. On the possible connections between Augustine’s ba-sic metaphysical arguments and Porphyry’s Mixed Questions and other works, seeJ. Pépin, Une nouvelle source de saint Augustin: le zÆthma de Porphyre « Sur l’union de l’âmeet du corps », «Rev. Ét. anc. », 66 (1964), 53-107.

Charles Brittain

SELF-KNOWLEDGE IN CICERO AND AUGUSTINE(DE TRINITATE, X, 5, 7-10, 16)

1. INTRODUCTION

In De trinitate, X Augustine argues that since the human mindnecessarily knows itself it must be an immaterial substance. Thebasic idea is a standard Platonist one about the nature of self-re-flexivity: psychological powers that depend on material substancescan’t be self-reflexive owing to the limitations of their materialparts. For sight to see itself in a genuinely self-reflexive way, onewould need to turn the eye round on itself, which would destroythe physical basis of sight.1 So, if the mind or intellect is essentiallyself-reflexive, as the Platonists think it is, it can’t depend on anymaterial parts. But, while the idea is a standard Platonist one, whatAugustine does with it in De trinitate, X, 5, 7-10, 16 is new. His orig-inality is captured in part in the two fairly well-known formalarguments that conclude his treatment in De trinitate, X, 10, 16(which I will call the “substance argument” and the “mode argu-ment’”). Perhaps more striking, however, is the self-reflexivemethod Augustine uses to try to make his reader grasp that she– or more properly, her mind – already knows herself (or, better,itself) in the relevant way.

In this essay, I will argue that Augustine achieved these resultsin De trinitate, X as a result of a deliberate reworking of an extend-ed argument in Cicero’s Tusculan Disputations, I, 22, 50-33, 81. I willtry to show that Augustine took from Cicero not just a doxogra-phy of ancient philosophical views about the nature of the mind(as scholars have recognized),2 but also the germ for his self-re-flexive method of argumentation; and that he employs these Ci-ceronian elements to show that the argument Cicero himself usesto indicate that we know some essential properties of the mind infact commits Cicero – and any materialist-leaning reader – to theview that it is an immaterial substance. The hope is that tracingout the Ciceronian context for Augustine’s argument will allow usto see more clearly what he is doing in De trinitate, X, and thus al-so illuminate its better-known formal arguments.

Augustine’s adoption of this self-reflexive method in De trinitate,X is easier to understand, I think, when we note that he has alreadyestablished its principal metaphysical claims in Book IX, includingthe basic conclusions that the mind is a substance (IX, 2, 232-33), thatit is immaterial (IX, 3, 316-19), and that it is a trinitarian structurewhose parts are not material but co-extensive with the whole andeach other (IX, 4, 7). These conclusions are reached in Book IX bya set of rather schematic metaphysical arguments which are likelyto leave a non-Platonist reader cold.3 And Augustine clearly recog-

2. See A. Schindler, Wort und Analogie in Augustins Trinitätslehre, J.C.B. Mohr,Tübingen 1965, Anhang I, 246-249, which is noted in H. Hagendahl, Augustine andthe Latin Classics, Almqvist & Wiksell, Stockholm 1967, I, 139-140; II, 514, andechoed, without much further comment, in most subsequent literature on trin., X,including P. Courcelle, Connais-toi toi-même de Socrate à Saint Bernard, I, Etudes Au-gustiniennes, Paris 1974-5, 120-122; R. Kany, Augustins Trinitätsdenken. Bilanz, Kritikund Weiterführung der modernen Forschung zu «De trinitate », Mohr Siebeck, Tübingen2007, 72-75, and L. Ayres, Augustine and the Trinity, Cambridge University Press,Cambridge 2010, 301. But see n. 10 below.

3. The arguments seem to be drawn from Porphyry’s Mixed Questions; seePépin, Une nouvelle source. The claim that the mind is a trinitarian substance, for ex-ample, is supported in IX, 4, 527-39 with the argument that the three activities can notbe “in a subject” like ordinary properties, because they “exceed” their subject, i.e.because the mind can know and love not just itself but also other things, whereasregular properties are confined to the bodies they are predicated of.

108 Charles Brittain

Stelf-Knowledge in Cicero and Augustine 109

4. See esp. trin., X, 12, 19, which explains why Augustine added Books XI-XIII toelucidate his main argument in Books X & XIV-XV; cf. trin., I, 1, 1-3, 6, esp. I, 3, 6;V, 1, 1-2; XI, 2, 3-4; XIV, 7, 10; XIV, 14, 20.

5. For treatments of trin., X as a whole, see M. Schmaus, Die psychologische Trinität-slehre des hl. Augustinus, Aschendorff, Münster 19672, 235-245; E. Bermon, Le « cogito»dans le pensée de Saint Augustin, Vrin, Paris 2001, 77-104; J. Brachtendorf, Die Struktur desmenschlichen Geistes nach Augustinus. Selbstreflexion und Erkenntnis Gottes in «De Trini-tate », F. Meiner, Hamburg 2000, 163-193, and Ayres, Augustine and the Trinity, 297-308.

6. Augustine in fact only clarifies this gradually as the work unfolds. The term(mens) is introduced in trin., IX, 2, 2; the empirical faculties (i.e. the ones requiring

nized that many of his readers were unlikely to see the force ofsuch arguments – he says as much in the contemporary Letter 169, 1& 3 and he refers frequently in De trinitate to the special devices hefinds it necessary to deploy in order to keep the “slower” part of hisaudience with him.4 The main function of Book X, then – or so Iwill argue – is to bring the ordinary reader up to speed. If so, theidea is not so much to prove that the mind is immaterial but to re-orient the reader so that her mind comes to see for itself what it is:the argument is a work-out or an exercise designed to transformthe reader by disclosing her mind’s necessary presence to itself.

A final caveat: Book X is structured in the form of a continuousargument in three parts (reflecting its status as a reworking ofBook IX):

Part I (X, 1, 1-4, 6) argues that the mind necessarily knows itself.

Part II (X, 5, 7-10, 16) argues that it accordingly knows itself as animmaterial substance.

Part III (X, 11, 17-12, 19) argues that its self-knowledge reveals itstrinitarian structure.

But in this essay, I address only Part II, the argument for themind’s immateriality based on its self-knowledge.5 To understandwhat is going on, it is thus useful to have in mind two basic pre-suppositions, which I will take for granted here. These are:

1. The term ‘mind’ (mens) in De trinitate refers to the higher ra-tional mind or intellect. This means that it excludes all empiri-cal capacities, acts, and dispositions, including perception, imag-ination, and sensory memory.6

2. The battery of (mainly) Platonist arguments in X, 1, 1-4, 6have shown that it doesn’t make sense to think that the minddoes not know itself completely – in the sense of knowing itsown nature or structure – since its nature as a knower presup-poses this.7

My aim here is to show how Augustine tries to make sense ofthis idea by entering into a dialogue with Cicero, the very modelof an educated reader who still needs to discover what Augustinethinks he necessarily knows.

2. CICERO

I start with Cicero’s argument for a limited form of self-knowl-edge of the rational soul in Tusculan Disputations, I. My aim in thissection is not to give a full account of this fascinating, but also con-fusing and under-studied text, which is best seen as Cicero’s re-sponse to the Plato’s Phaedo.8 What I want to show is just that Au-gustine rightly identified and took over three elements of the ex-

representation) are ruled out explicitly in X, 7, 10; it is defined as a “rational intelli-gence” in X, 9, 13; and we discover in XII, 1, 1-3 that even the rational use of repre-sentations is non-essential to it; see C. Brittain, Intellectual Self-Knowledge in Augus-tine («De trinitate » 14.7-14), in E. Bermon - G. O’Daly (cur.), Le De trinitate de saintAugustin: exégese, logique et noétique. Actes du colloque international de Bordeaux, 16-19 juin2010, Institut d’Etudes Augustiniennes, Paris 2012, 321-339, at 321-326.

7. These arguments will be the subject of a parallel paper on Augustine’s use ofPlotinus in trin., X & XIV. The best recent treatments are C. Horn, Selbstbezüglichkeitdes Geistes bei Plotin und Augustinus, in J. Brachtendorf (cur.), Gott und sein Bild. Au-gustins «De Trinitate» im Spiegel gegenwärtiger Forschung, F. Schöningh, Paderborn 2000,81-103, and two of Jean Pépin’s last works on Augustinian themes: Le tout et les par-ties dans la connaissance de la «mens » par elle-même (De Trin. X 3, 5-4, 6). (Nouveauxschèmes porphyriens chez Saint Augustin, III), in Brachtendorf (cur.), Gott und sein Bild,105-126, and Porphyre et Plotin sur le tout et les parties dans la connaissance de l’Intellect parlui-même, in M. Barbanti - G.R. Giardina - P. Manganaro (cur.), HENOSIS KAI PHILIA =Unione e amicizia: omaggio a Francesco Romano, CUECM, Catania 2002, 399-407.

8. Though see I. Gildenhard, Paideia Romana: Cicero’s «Tusculan Disputations »,The Cambridge Philological Society, Cambridge 2007, 207-255 for a fine start at se-rious literary analysis of Book I. (Explicit references to the Phaedo in Tusc. disp., I arelisted in n. 9 below.)



9. See E. Emilsson, Platonic Soul-Body Dualism in the Early Centuries of the Empire toPlotinus, in W. Haase - H. Temporini (cur.), Aufstieg und Niedergang der römischen Welt,II.36.7, W. de Gruyter, Berlin-New York 1994, 5331-5362, who (rightly) takes Tusc.disp., I as a useful summary of Platonist psychology at the start of the revival of Pla-tonism in the 1st Century BC. Cicero himself identifies the arguments for immor-tality he gives in Tusc. disp., I, 22, 50-33, 81 as based on Plato but also drawing on anearlier Pythagorean and Italian tradition; see Tusc. disp., I, 16, 38-17, 39 & I, 21, 49. Pla-to and specific Platonic dialogues are cited in I, 24, 57-58; I, 28, 63; and I, 29, 71-31, 75is a sort of paraphrase of stretches of the Phaedo.

10. See Aug., trin., X, 7, 9, which is based directly on Cic., Tusc. disp., I, 9, 18-10, 22(cf. n. 2 above). Partial exceptions are P. Cary, Augustine’s Invention of the Inner Self,Oxford University Press, New York 2000; S. Menn, Descartes and Augustine, Cam-bridge University Press, Cambridge 1998 – see n. 13 below – and, most notably,Bermon, Le « cogito », 84-87, whose remarks there on the parallels between Ciceroand Augustine anticipate points [a]-[c] above.

plicitly “Platonist” part of the argument in Tusculan Disputations, I,22, 50-33, 81.9 The three elements are:

[a] his main argument for the immortality of the soul based on itsself-knowledge, which relies on

[b] an extended, if sometimes implicit, appeal to a set of subjectiveand performative facts about the participants in the inquiry(Cicero and his interlocutor),

and is bolstered by

[c] an error theory explaining his opponents’ inability to grasp therelevant facts.

The complicated structure of Tusculan Disputations, I makes iteasy to see why patristic scholars have often noticed that Augus-tine used it as a doxographical source for philosophical disagree-ments about the nature of the soul, but have consistently failed toidentify his reading – and re-application – of Cicero’s method andargument.10 So let me note a few points about the structure of Tus-culan Disputations, I. Its general aim is to persuade the unnamed in-terlocutor that death is not bad by means of an over-arching dis-junctive argument that if the soul is mortal, death is not bad, andif the soul is immortal, death is actually good. Cicero treats thesecond disjunct in I, 9, 17-33, 81, which thus give a series of argu-ments for the immortality of the soul. The Platonist argument in

this series (I, 22, 50-33, 81) – the one we are concerned with now –comes after a doxography of views about the nature of the soul(I, 9, 18-10, 21), a set of arguments for immortality from consensus(I, 12, 26-16, 37), and some further immortality arguments based oncontemporary Stoic physics (I, 17, 40-21, 49).

Since the main argument is set out over a long stretch of textand is itself an over-arching argument containing internal disjunc-tive premises (which I don’t examine here),11 it will be useful toset it out schematically, I think, before we review Cicero’s proce-dure in the text.

The argument is something like this:

1. There is an apparently intractable dispute about the nature ofsoul, i.e. about whether it is fire or air or immaterial etc.

2. But all parties ought to agree that at least one set of properties isessential to the soul, namely the set of properties undeniablymanifested by any agent in the course of this inquiry – viz. act-ing, thinking, remembering, etc.

3. This set of essential properties is “divine” because it parallelsthe set of properties we ascribe to god as the mind who struc-tures and orders the world.

4. Hence, if god is immortal, so is the soul. (Or, more plausibly:hence, the grounds for our thinking that god is immortal are al-so grounds for our thinking that the soul is immortal.)

Note that Cicero’s conclusion is an inference to the soul’s im-mortality on the basis of its limited self-knowledge: the argumentis that even if we don’t know what the nature of the soul is, we doknow enough of its essential properties to infer that it is god-likeand so immortal.

11. Cicero gives several different versions of the “diuina” argument sketched be-low, depending on different views about the nature of the soul, viz. (a) what seemshis current view, that it is either fire or air (Stoic views) or a quintum quid (“Aristo-tle’s view”), or (b) that is something sui generis (a different version of “Aristotle’sview”).



12. Cic., Tusc. disp., I, 16, 37-38: « animos enim per se ipsos uiuentis non poterantmente complecti, formam aliquam figuramque quaerebant [...] nihil enim animouidere poterant, ad oculos omnia referebant. [38] Magni autem est ingenii seuocarementem a sensibus et cogitationem ab consuetudine abducere » (cf. I, 22, 50).

Cicero introduces this main argument, though, with an errortheory ([c] above), designed to dispel some empiricist misconcep-tions about the soul’s survival. The theory follows from the set ofarguments for the immortality of the soul from consensus in I, 12,26-16, 37, which show that everyone is naturally disposed to be-lieve that the soul survives. But, Cicero explains:

<ordinary people and the poets> weren’t able to understand the notion ofsouls living by themselves [...] they were unable to see anything with theirminds (mente), and referred everything to their eyes. It takes a remarkablecharacter to disassociate one’s mind from the senses and retract one’sthought from habit.12

In I, 22, 50-51, Cicero uses this critique of empiricism to explainwhy the Epicureans – the main philosophical opponents of theimmortality of the soul – are driven to this unnatural position.The Epicureans go wrong, he argues, owing to the same concep-tual poverty found in the poets they are so quick to deride. Theyare right to reject the poetic tales of anthropomorphic souls in theafterlife, which presuppose the survival of the body; but they arewrong to infer that the soul is therefore mortal. For this wouldonly follow if the soul’s form implied that it couldn’t survive on itsown. But we have no reason to believe that, since the sort of em-pirical evidence the Epicureans restrict themselves to can’t evenshow what its form or shape is while it is in the body, let alone af-ter death. So, Cicero concludes, we should not infer from our lackof empirical evidence about its shape that we cannot grasp the na-ture or fate of the soul with our minds (I, 22, 51).

For a reader of Augustine, there are two things to note aboutthis error theory. The first is that although it serves as an introduc-tion to the main Platonist argument in Tusculan Disputations, I, theargument is not itself a Platonic or Platonist one. This is clear ifonly because the same argument is found in Cicero’s earlier

13. See Aug., conf., VII, 17, 23 (& ep., 137, 2, 5), discussed in Cary, Augustine’s Inven-tion, 80-85 and in Menn, Descartes and Augustine, 139 n. 9.

14. Cic., Tusc. disp., I, 22, 52: «Non enim, credo, id praecipit, ut membra nostraaut staturam figuramue noscamus, neque nos corpora sumus, nec ego tibi haec di-cens corpori tuo dico. cum igitur ‘nosce te’ dicit, hoc dicit: ‘nosce animum tuum’.Nam corpus quidem quasi uas est aut aliquod animi receptaculum: ab animo tuoquicquid agitur, id agitur a te ».


De natura deorum, II, 17, 45 (cf. III, 8, 20-21), where it is presented asa Stoic argument against Epicurean anthropomorphism aboutgods. The second is that, while it is easy to see how Augustinecould adapt it for use against materialist thinking as such – as hedoes in Confessions, VII with a verbal allusion to Tusculan Disputa-tions, I, 17, 38, as well as in De trinitate, X13 – in Cicero’s texts it doesnot indicate that the soul is immaterial. This error-theory is an ar-gument against « those who say that they cannot conceive of thesoul without a body » (Tusc. disp., I, 22, 51), but the kind of concep-tual poverty it is aimed at is not materialism as such, but rather ananthropomorphism based on excessive reliance on empiricalmodes of thought.

So much for the error theory ([c] above). With this epistemo-logical introduction in place, Cicero turns to the main argumentin I, 22, 52 with an appeal to the Delphic Oracle:

I don’t believe that the Delphic precept is telling us to know our limbs,stature or shape: we are not our bodies, nor am I talking to your body asI speak. So when it says “Know yourself!” it means: “Know your soul!”For the body is a sort of vessel or container for the soul; but whatever isdone by your soul is done by you.14

The vital point here for the argument that follows is not the So-cratic identification of the soul as “you” or your self, which Cicerotreats as straightforward, but rather the way in which you are sin-gled out as an agent: what you or your soul are is a locus of agency.The next twenty sections set out the main argument by examin-ing the kinds of agency the soul is capable of, i.e. by looking at itscapacities. He identifies three sorts of psychological capacity:

i) the soul’s capacity to move, based on its own self-motion(I, 22, 53-23, 55);


15. Thus in Tusc. disp., I, 25, 60 on memory, for example, Cicero argues that wedon’t fully understand the nature of memory, but we can infer from its psycholog-ical power that it is not derived from heavy physical elements, but rather “divine”.The general form of these “diuina” arguments is probably derived from Aristotle’s(lost) De Philosophia, but Cicero’s deployment of it is sui generis, like his implausiblecharacterization of Aristotle’s views on the soul in the doxography in Tusc. disp.,I, 10, 22; see H. Easterling, Quinta Natura, «Mus. Helvet. », 21 (1964), 73-85;D. Halm, The fifth element in Aristotle’s «De philosophia », « J. Hellen. Stud. », 102(1982), 60-74, and D. Furley, Aristotelian Material in Cicero’s «De natura deorum»,in W.W. Fortenbaugh - P. Steinmetz (cur.), Cicero’s Knowledge of the Peripatos, Trans-action Publishers, New Brunswick (NJ) 1989, 201-219, at 205-211 (which discussedthe Ciceronian “endelecheia”).

ii) its capacity to cause life, through desire (I, 24, 56);

iii) its higher cognitive capacities, including especially memoryand thought (I, 24, 57-26, 65).

In each case, his procedure is to examine the capacity, with ex-plicit reference to celebrated arguments from Plato, in order tosee what sort of cause we should ascribe the capacity to. In thethird case in particular, he draws the conclusion that I have set outabove in my initial sketch of the main argument: while we maynot be in a position to know precisely what the nature of the soulmust be to have the relevant properties, we do know enough tosee that these are “divine” properties.15

To see exactly what Cicero is doing, we would need to see nexthow he moves from the set of essential psychological propertieshe identifies as “divine” to the inference he draws from it, invirtue of the soul’s similarity to god, to the immortality of thesoul. But since our interest in Cicero here is directed towards un-derstanding Augustine’s reading of him, I am going to ignore hisformal conclusion and instead focus on some special features ofhis arguments that the soul has these “divine” properties.

The features I have in mind are the elements of subjectivity andperformativity I singled out above as the third element of Cicero’sargument recognized and adapted by Augustine ([b] above). Thisthread in the presentation is present, I think – at least implicitly –in the construction of the argument as outlined above and startingwith the Delphic precept in Tusculan Disputations, I, 22, 52 (cited inn. 14 above). For the idea is that we can find immediate and (rela-

tively) indubitable premises about some of the essential propertiesof the soul – if perhaps not enough to disclose its whole nature –by looking at its psychological agency. And, unlike its material ormetaphysical composition, the soul’s psychological agency issomething that is immediately given to it in the very act of inquiryor argument. If something like this is on the right lines, a majorstrand of Cicero’s argument appeals to three features of psycho-logical agency:

i) subjectivity: the fact that it is you, your own case, that is in ques-tion, is central to the success of the arguments for the soul’sagency.

ii) performativity: the fact that your activity in inquiring or argu-ing in this way itself establishes or warrants your possession ofthese psychological properties.

iii) mode of presentation: the fact that the relevant psychologicalactivities are known by you immediately in virtue of your beingtheir subject or agent.

In my view, the arguments Cicero gives in Tusculan Disputa-tions, I, 22, 50-33, 81 exhibit clear indications of these three fea-tures. I, 22, 52 sets us up to expect them by appealing to “you”.(This is the “you” of self-reflexivity, rather than the generic “you”or “one”, since the point of the Delphic precept is that you knowyourself as an agent, not just that you know that human souls areagents.) Something of the same kind can be found in the first caseof psychological agency Cicero gives, the case of the soul’s self-motion in I, 22, 53-23, 55, which is set out explicitly in terms of thefamous Platonic argument for the immortality of the soul inPhaedrus, 245 C-E. Without attempting to consider the complexi-ties and success of the argument that the soul is a self-mover andhence an immortal principle in virtue of its role as a cause of mo-tion in other things, we can still note a peculiarity of his presenta-tion of this argument.16 For Cicero concludes his translation witha summary not found in the Platonic original:

16. Cicero’s translation is important, inter alia, for the implied text of Phaedrus,245-246, and for its application of the argument to individual souls; see K. Büchner,


So the soul feels that it is moved; and when it feels that, it also feels that itis moved by its own rather than an external force – and that it isn’t possi-ble that it should ever be abandoned by itself.17

Cicero’s point here (at least in the first part of the sentence) is,I think, to stress the direct application of the self-motion argu-ment to one’s own case: going through the argument is itself acase of self-motion, i.e. an example of psychological activity. As aresult, it is something that Cicero or the interlocutor or the readerhas immediate access to – it is something that is warranted and feltor experienced directly by performing it. (One reason to thinkthat something like this is going on here in Tusculan Disputations, I

– in this and in the following “diuina” arguments treated below –is that Cicero gave the same arguments in De re publica, VI, 24, 26-26, 28 and De senectute, 21, 77-22, 79 without any suggestion of self-reflexivity.)

But the clearest evidence for seeing the presence of an appeal tosubjectivity etc. in Tusculan Disputations, I is the central case of thesoul’s capacities for memory and thought in I, 24, 57-26, 65. Again,Cicero does not say it explicitly, but his long excursus on the pow-er of memory is of course itself an example of it: as he explains atlength in his rhetorical treatises, the training of our memories is asignificant part of the discipline of rhetoric – so his speech here inTusculan Disputations, I is meant to be a particularly brilliant instan-tiation of his own memorial skills.18 The speech is also, of course,an instantiation of “thought”, which includes the ability to makephilosophical arguments, as he argues in I, 26, 64. Cicero’s reliance

M. Tullius Cicero. De Re Publica, C. Winter, Heidelberg 1984, 502-504 and R. Bett,Immortality and the Nature of the Soul in the Phaedrus, in G. Fine (cur.), Plato, 2, Ethics,Politics, Religion, and the Soul, Oxford University Press, Oxford 1999, 245-249.On Augustine’s use of this argument in Cicero, see Bermon, Le « cogito », 366-369.

17. Cic., Tusc. disp., I, 23, 55: « Sentit igitur animus se moueri; quod cum sentit, il-lud una sentit, se ui sua, non aliena moueri, nec accidere posse ut ipse umquam ase deseratur ».

18. See his conclusion in I, 25, 60 on the “extension” of memory: the interlocu-tor can at least see how “big” Cicero’s memory is from his amazing display oflearning in Tusc. disp., I.


on the self-reflexivity of his argument here is plain, I think, fromhis general conclusion to the central case in I, 27, 67:

“So where is this mind (mens) and what sort of thing is it?” Well, where isyours and what is it like? Can you say? If I’m not in a position to knoweverything I’d like to, is it ok by you if I use what I am in a position toknow? “The soul is not strong enough to see itself, but like the eyes, thesoul discerns other things without seeing itself.” Well, it doesn’t see itsform, the least important thing about it (though perhaps it can see thattoo, but let’s not get into that now). But it surely does see its force, itssagacity, its memory, its motion, its speed. These are great things, divineand eternal things – whereas its form and location aren’t even worth in-vestigating.19

Cicero is too polite to point out explicitly that the force, sagaci-ty, memory, <psychological> motion and speed of his own soul isimmediately and indubitably known to him as he makes this ar-gument. But it is precisely the particular soul that is doing thesethings that can see these essential properties as things that belongto it just because it is doing them. And that’s why he points out tothe imaginary interlocutor here that she should look at her mind– what he can say is that his soul has at least these properties.20

That last point is re-iterated, I think, in the final argument ofthe “main” Platonist part of Tusculan Disputations, I, where Ciceroconcludes:

19. Cic., Tusc. disp., I, 27, 67: «Vbi igitur aut qualis est ista mens? – Vbi tua autqualis? potesne dicere? an, si omnia ad intellegendum non habeo, quae habereuellem, ne iis quidem, quae habeo, mihi per te uti licebit? – Non ualet tantum a-nimus, ut se ipse uideat. At ut oculus, sic animus se non uidens, alia cernit. – Nonuidet autem, quod minimum est, formam suam (quamquam fortasse id quoque,sed relinquamus); uim certe, sagacitatem, memoriam, motum, celeritatem uidet.Haec magna, haec diuina, haec sempiterna sunt; qua facie quidem sit aut ubihabitet, ne quaerendum quidem est ». (Dougan starts the second interjection, inline 3, at “at”; Pohlenz does not mark it at all.)

20. Note that the switch from animus (“soul”) to mens (“mind”) here is renderedinnocuous by Cicero’s remark in Tusc. disp., I, 33, 80 that the arguments he has beendiscussing are properly construed as arguments about the mind rather than the soul,viz. as arguments about rational (and so non-emotional and non-passive) soul.


But let’s deal with the location of the soul another time; it’s at least clearthat it’s in you. So what is its nature? Something unique to it, I think. Butassume it’s fiery or airy – it won’t make any difference for our purpose.Just make sure that you see that, just as you know god although you don’tknow his location or form, so your soul should be known to you, even ifyou don’t know its location or form.21

This conclusion relies, admittedly, on the parallel between thesoul and god that grounds his inference from the “divine” proper-ties of the soul to its immortality. But I don’t think that the factthat Cicero can only establish this parallel by indirect evidence inthe case of god (we infer god’s nature from the order in the world)undermines my suggestion that Cicero is appealing to subjectiveand performative facts in the case of the soul. For the source of hiscertainty about the limited set of psychological properties just isthe fact that they are immediately given in the act of inquiry.

So much then for Cicero. I hope to have shown that in TusculanDisputations, I, 22, 50-33, 81 he offers the “main argument” backedup by an anti-empiricist error theory and at least some hints of amethod that relies crucially on the appeal to subjectivity.

3. AUGUSTINE’S USE OF CICERO

In this section I will try to show that Augustine saw these threefeatures in Cicero’s argument in Tusculan Disputations, I and adapt-ed them to his own ends in De trinitate, X, 5, 7-10, 16. But my aim isto show not just that we can find a similar structure in both texts –viz. an argument about the nature of the mind based on its self-knowledge, introduced by an error theory and supported by a par-ticular kind of appeal to subjectivity – although that would alreadybe an interesting result, I think. Rather, as I said at the start, myhope is to show that it is by seeing Augustine’s procedure as a

21. Cic., Tusc. disp., I, 29, 70: «Sed alias, ubi sit animus; certe quidem in te est. Quaeest ei natura? Propria, puto, et sua. Sed fac igneam, fac spirabilem: nihil ad id de quo a-gimus. Illud modo uideto, ut deum noris, etsi eius ignores et locum et faciem, sic ani-mum tibi tuum notum esse oportere, etiamsi ignores et locum et formam».


reading of Cicero and a dialogue with him, that we can best ap-preciate his arguments in De trinitate, X.22

The structural relation between the two texts is interesting,though, in any case, because the two authors have different aimsand different starting points. Augustine – unlike Cicero – starts inX, 5, 7 with a conclusion he has already established: the mindknows itself permanently and as a whole, i.e. entirely. As a result,his aim in X, 5, 7-10, 16 is not so much to establish the nature of themind or one or some of its essential properties as to show the read-er how it is – or rather that it is the case – that she already has therelevant kind of self-knowledge. The formal function of the mainargument (and the two formal arguments Augustine concludeswith) is to show that the essential properties of the mind presup-posed by its act of self-inquiry in fact exhaust its essential properties:the nature of the mind just is to be a conscious subject. But thepurpose of the passage is not so much the affirmation of conclu-sions Augustine has been assuming throughout De trinitate, and ar-gued for already in Book IX, but rather to disclose her own mind tothe reader, or to get the reader into a position to see what, accord-ing to Augustine, she already knows.23 In what follows, then, I willtry to show how Augustine does this in three stages, correspondingto our three Ciceronian features: [c] by setting out an error theoryin X, 5, 7-7, 10, [b] by a series of subjective and performative exercis-

22. The history of Augustine’s multiple and sophisticated readings of Cicero re-mains largely unwritten, despite the early start to the topic given by M. Testard,Saint Augustin et Cicéron, Etudes Augustiniennes, Paris 1958. One complication isthat Augustine saw Cicero as a philosophically ambivalent figure, an Academicand Platonist; see C. Brittain, Augustine as a reader of Cicero, in R. Taylor - D. Tweet-en - M. Wreen (cur.), Tolle Lege. Essays on Augustine & on Medieval Philosophy in Hon-or of Roland J. Teske, SJ, Marquette University Press, Milwaukee 2011, 81-114.

23. To this extent I agree with the revisionist theological readings of trin., X inR. Williams, The Paradoxes of Self-Knowledge in the «De trinitate », in J.T. Lienhard -E.C. Muller - R.J. Teske (cur.), Augustine: Presbyter factus sum. Collectanea Augustiniana,P. Lang, New York 1993, 121-134, and L. Ayres, The Discipline of Self-knowledge in Au-gustine’s «De trinitate » Book X, in L. Ayers (cur.), The Passionate Intellect: Essays on thetrasfrmation of classical tradition, presented to professor I.G. Kidd, Transaction Publishers,New Bruswick (NJ) - London 1995, 261-96. But, as I try to show below, the disciplineor training of the reader is achieved by careful and explicit philosophical argument.


es to orient the reader in X, 7, 10-9, 13, and [a] by a revision of theCiceronian main argument aimed at persuading someone like Cic-ero – an educated but open-minded reader – that his mind really isan immaterial substance, and he knows it, in X, 10, 13-16.

3.1. The error theory

Augustine starts with an error theory in De trinitate, X, 5, 7-7, 10,because his conclusion in X, 4, 6 that the mind knows itself as awhole is perplexing:

So what’s the point of the mind being told to know itself?24

The problem is not just that the Delphic precept seems otiose,since Augustine has an immediate answer to that:

Presumably so that it should think itself and live in accordance with itsown nature [as opposed to just knowing itself dispositionally].25

The problem is that, as Augustine sees it, most people who dothink about their minds (including most Christians and a goodnumber of philosophers) don’t think of them properly. Most peo-ple, after all, are materialists, so whether or not they accept anygiven philosophical view of the mind, they think of it is as if itwere a material thing of some kind.

Accordingly, Augustine proposes an error theory to explain howit is that so many people appear to hold views that are incompati-ble with his conclusion that the human mind necessarily knows it-self. The theory has two strands. The first is a standard Platonist ex-planation of materialism as the result of the mind’s excessive inter-est in its own cognitive and desiderative activities and their exter-nal objects. (The theory is adumbrated in Plato’s Phaedo and set outmore fully in texts such as Plotinus, Enneads, V, 1, 1-3 [which is veryclose to De trinitate, X, 5, 7-7, 10].) The idea is that human beings be-

24. Aug., trin., X, 5, 7, init. (CCL 50, 320): «Vtquid ergo ei praeceptum est ut se ip-sa cognoscat? ».

25. Aug., trin., X, 5, 7, cont. (CCL 50, 320): «Credo ut se cogitet et secundum na-turam suam uiuat ».


come so accustomed to thinking and dealing only with materialobjects that they are no longer able to conceive of anything imma-terial. But, Augustine argues – and this is the second strand – evenin its continual dealings with the material world and its own body,the mind can’t be fully occluded from itself. So the problem is notthat their minds aren’t present or given in their mental experience– they remain the subjects of their own thoughts – but rather thattheir conceptual poverty makes them represent their minds tothemselves as something material, albeit (at least tacitly) still some-thing with the mental properties that can’t be occluded.

Augustine sums up the problem at the conclusion of his reviewof philosophical views about the nature of the mind (which isdrawn directly and almost verbatim from Cicero’s doxography inTusculan Disputations, I, 9, 18-10, 22):

Amidst all these views, anyone who sees that the nature of the mind isboth a substance and non-material (that is, something whose parts do notoccupy more or less space in proportion to their size) should also then seethat those who take it to be material are not in error because the mind isnot present to their understanding but because they add to it the thingswithout which they can’t think any nature. For they believe that anythingthey are asked to think of without material representations is nothing atall. So the mind should not seek itself as if it is not present to itself.26

3.2.The exercises

The error theory thus gives a diagnosis of the materialist’sand ordinary reader’s condition. The next stage – in De trinitateX, 7, 10-9, 13 – is a set of warm-up exercises designed to preparethe reader to see why she is already in a position to reject mate-rialism about the mind. These exercises follow the prescription

26. Aug., trin., X, 7, 1034-42 (CCL 50, 323): « In his omnibus sententiis quisquis uidetmentis naturam et esse substantiam et non esse corpoream, id est non minore suiparte minus occupare loci spatium maiusque maiore, simul oportet uideat eos quiopinantur esse corpoream non ob hoc errare quod mens desit eorum notitiae, sedquod adiungunt ea sine quibus nullam possunt cogitare naturam; sine phantasiisenim corporum quidquid iussi fuerint cogitare nihil omnino esse arbitrantur;ideoque non se tamquam sibi desit mens requirat ».


at the end of the passage cited above: they show us how to iso-late the mind as such from the material accretions we have con-flated it with, by showing our minds how they are necessarilypresent to themselves in the act of self-inquiry.

Augustine offers three such exercises in X, 7, 10-9, 13.27 Theseexercises are an invitation to think about the difference betweenthe subject of any normal inquiry and its intentional object, on theone hand, and the case of the mind’s self-inquiry, on the other. First(X, 7, 1042 - X, 8, 1113), when the mind looks for something else it di-rects its attention outside on to the object, and we say that the objectis now “in the mind” when the mind successfully finds it or focuseson it.28 But in the case of self-inquiry, this doesn’t make sense: themind is already “in” itself in the second sense, because it is the sub-ject of the inquiry. So when it’s told to “Know itself!” it needs to getaway from what is alien to it, not find something beyond itself. Sec-ondly (X, 8, 1113-28), that means that it should be looking inside itself– and not just inside as far as its representations of material things,but “right inside”, at itself as the subject of these thoughts.29

What Augustine is doing with these first two exercises is, I sug-gest, an intensified version of Cicero’s appeals to subjectivity inTusculan Disputations, I. Where Cicero just intimated the self-reflex-ive features of his argument – the fact that his discussion of thepower of memory and thought presupposes their exercise in his

27. On these exercises, see Bermon, Le « cogito », 229-239, and L. Hölscher, TheReality of the Mind: Augustine’s philosophical arguments for the human soul as a spiritual sub-stance, Routledge & Keagan Paul, London 1986, 128-142.

28. For the idea that the mind finds “things” in different parts of itself, cf. trin.,IX, 3, 3. Augustine construes this in terms of inuentio (“discovery” or “coming uponsomething”) here partly to pick up on the original designation of the problem ontrin., X at the end of trin., IX, 12, 19, where the mind’s self inquiry is defined as an in-tense form of a desire for the discovery of, or coming upon, something.

29. Augustine characterizes the rational aspect of the human mind – i.e. the« part » animals don’t have – as the « inner man»; see trin., IV, 2, 6; XI, 1, 1; XII, 1, 1-2.Hence he often characterizes imagination as further « inside » (interior) than per-ception – see trin., XI, 4, 7; XI, 7, 11 & XV, 9, 18 – and talks of the mind’s intellectualcore as it’s actual « inside »; see trin., X, 7, 11 (here) & X, 10, 16. (In trin., XIV, 7, 10 welearn there is an innermost core of the intellect, which consists in its constituitiveself-knowing, self-remembering and self-loving.)


own case, etc. – Augustine has made it central: the exercises are ex-plicitly what the mind – a mind, your mind – should do whenseeking itself. In the third exercise (X, 9, 121-1325), the element ofperformativity in these exercises is also made explicit. For in thiscase, the exercise is simply listening to the imperative “Know your-self!”: Augustine claims that anyone who understands the meaningof these words (so any competent language speaker) immediatelyknows herself on hearing this command. The positive claim he isurging on us here is that the mind is necessarily “present” to itself.30

But Augustine presents this case by imagining an explicit per-formance or series of mental actions in response to imperatives:

when the mind is told to “Know yourself!” it’s not the same as its beingtold to “Know the cherubim and seraphim!” since in their case we formbeliefs about them as things that are not present, in accordance with theirstatus in prophecy as heavenly powers of some sort. Nor is it the same asits being told to “Know the will of that (human) person!” which is inprinciple unavailable to perception and only available to our understand-ing as a result of a display of bodily signs, and accordingly <only> in sucha way that it is more a case of believing than understanding. Nor is it thesame as a person being told to “Look at your face!” which can’t be doneexcept in a mirror, because our own face is not present to our gaze eithersince it is not in a place to which the latter can be directed.31

This sequence of imperatives offers a negative argument bycontrasting the special case of self-knowledge with the mind’s ac-cess to a list of other kinds of intentional object. The force of theargument derives from the theory of contents Augustine outlinedin Books VIII-IX and the earlier part of X (X, 1, 1-4, 6) and goes on

30. On the mind’s necessary and direct or unmediated presence to itself, seetrin., X, 2, 5; X, 9, 12 (cited in part in n. 31 below); X, 10, 16 & XIV, 11, 14.

31. Aug., trin., X, 9, 126-16 (CCL 50, 325-326): « quia non ita dicitur menti‘Cognosce te ipsam’ sicut dicitur ‘Cognosce cherubim et seraphim’: de absentibusenim illis credimus secundum quod caelestes quaedam potestates esse praedican-tur. Neque sicut dicitur ‘Cognosce uoluntatem illius hominis’ quae nobis nec adsentiendum ullo modo nec ad intellegendum praesto est nisi corporalibus signiseditis, et hoc ita ut magis credamus quam intellegamus. Neque ita ut dicitur homi-ni ‘Vide faciem tuam’, quod nisi in speculo fieri non potest. Nam et ipsa nostra fa-cies absens ab aspectu nostro est quia non ibi est quo ille dirigi potest ».


to fill out in Books XI-XV. Using this theory we can see that Au-gustine is lining up a range of intentional objects with the variousmodes of presentation he has identified in order to point to theways in which they can be present to the mind. So:

1. Cherubim and Seraphim are things we can only think ofthrough imagination, via what Augustine calls a phantasma, be-cause they are not present to the mind at all (we know of themonly by authority).32

2. The will of another person – in the sense of what they want –is something we can only have in mind of through a represen-tation, via what Augustine calls a phantasia, because it is presentto the mind only indirectly through perceptible signs (such aswords or gestures).33

3. One’s own face is something we can see through perception, buteven in that case, it is still present to the mind only indirectly, be-cause it is available only through a reflected representation.34

The case of the mind’s access to itself, however, is quite different:

But when the mind is told to “Know yourself!”, it knows itself at the verymoment when it understands the expression ‘yourself ’ and it does so forno other reason than that it is present to itself. Whereas if it doesn’t un-derstand what was said, it obviously doesn’t do it. Thus it is commandedto do just what it does when it understands the command itself. [13]Hence, it should not conjoin anything else to what it knows as itself whenit hears that it should know itself. After all, it knows for sure that <this> isbeing told to it, that is to the <mind> itself that exists and is alive and un-derstands. Yet corpses also exist, and beasts are also alive: but neithercorpses nor beasts have understanding. Hence it knows that it exists andis alive in the way an intelligence exists and is alive.35

32. Cf. Aug., conf., XII, 22, 31; the authority is Isaiah 37, 6 (Cherubim) & 6, 2(Seraphim).

33. This case is one Augustine considered first in trin., VIII, 3, 4 and especially IX,6, 11, and goes on to explore in more detail in XIII, 1, 3 and XIII, 2, 5.

34. Cf. n. 1 above.35. Aug., trin., X, 9, 12-1316-26 (CCL 50, 326): « Sed cum dicitur menti : ‘Cognosce

te ipsam’, eo ictu quo intellegit quod dictum est te ipsam cognoscit se ipsam, nec


Augustine draws four conclusions here, I think. The first isthat the mind’s access to itself is immediate and direct because itis necessarily present to itself, as he has been arguing all throughX, 7, 10-9, 13. The second is that it is not just available to itself dis-positionally, but actively and occurrently known to itself whenany competent language speaker hears the imperative “Knowyourself!”. The import of this conclusion is clarified, I think, bythe third one, that, as a result, the mind is available for completeself-knowledge or conscious self-thinking immediately when itsattention is directed just at itself without any accretions (i.e. with-out representing itself as something more than it is). For if weconsider what might be given to all English-speakers when theyhear the command “Know yourself!” no matter what ideas theymay have about the nature of the mind – this is the case the sec-ond conclusion concerns – it seems plausible to suggest that it issomething like the notion of being a phenomenological subjectof thought or being the conscious thinker who is thinking. If so,the third conclusion is that this is precisely what the mind is itself,when one thinks just it – i.e. the whole of it but without any alienadditions. And the fourth conclusion, given in the last three sen-tences in this passage, is that this is what the mind is because themind’s life and existence consists in the activity of understanding,i.e. being a conscious subject of thoughts.

At this point, Augustine has finished with the warm-up exercis-es and is ready to introduce the main argument (De trinitate, X, 10,13-16). But before turning to that, it will be useful, I think, to notetwo questions for Augustine’s account so far. Let us grant Augus-tine that the sort of performative exercise he has just given us is ad-equate to isolate and secure the conclusion that being a phenome-nological subject is an essential property (or set of properties) of

ob aliud quam eo quod sibi praesens est. Si autem quod dictum est non intellegit,non utique facit. Hoc igitur ei praecipitur ut faciat quod cum praeceptum ipsumintellegit facit. [13] Non ergo adiungat aliud ad id quod se ipsam cognoscit cum au-dit ut se ipsam cognoscat. Certe enim nouit sibi dici, sibi scilicet quae est et uiuit etintellegit. Sed est et cadauer, uiuit et pecus; intellegit autem nec cadauer nec pecus.Sic ergo se esse et uiuere scit quomodo est et uiuit intellegentia ».


the mind. Even so, it is not yet clear why we should also agree thatthat property exhausts the mind’s essential properties, i.e. that thatis the nature of the mind full stop – which is the conclusion of X,4, 6 and what X, 5, 7-10, 16 is designed to make us see. This is hardto accept, I think, even when we recall that Augustine has explicit-ly limited the meaning of ‘mind’ in De trinitate, X to the highermind or intellect, and thus excluded any capacities, dispositions oracts that involve representations – and so has excluded perceptionand imagination along with empirical thought and memory, etc.For even so, it is hard not to think that the intellect has certain ca-pacities etc. that aren’t given in conscious thought or captured bythe idea that it is the subject – for instance, its dispositional mem-ory of intellectual objects or its capacity to make judgments (as op-posed to its conscious acts).

The second question we should keep in mind, I think, concernsthe way Augustine has construed the exercises and hence theiractual results in performance. At issue here is his insistence inDe tri-nitate, X, 5, 7-10, 16 on addressing the “mind” and telling “it” to thinkor act in a certain way. (The entire passage is written in the thirdperson singular and in the imperative mode.)36 The question iswhether any commands addressed to the mind can be performed atall – or if they can, whether their results are as clear as Augustinesuggests – given that it is not my or the reader’s mind that reads thetext or hears the command, but me or the reader. When I hear thecommand “Know yourself!” I understand it in virtue of a set of cog-nitive activities that necessarily include perceptual representations.It is not so clear that the end result is a distinct purely rational or in-tellectual activity. (To put this another way, this is one you can try athome. What does happen when I say: “Know yourself!”?)

These are both questions or problems that Augustine wasaware of, since he alludes to them directly or indirectly in the lat-

36. Augustine uses third singular addresses to the mind throughout through outtrin., X, 5, 7-6, 8; he switches to third plural reports of the views of the materialistsin X, 7, 9, but reverts to third singular in X, 7, 10-9, 13, until the materialists re-enteras interlocutors in X, 10, 13-16.


er books of De trinitate, so they may be ones his fuller account cansolve.37 But they indicate two points where it might be reasonableto question Augustine’s revision of Cicero’s approach.

3.3. The main argument

After the error theory and exercises in De trinitate, X, 5, 7-10, 13,the reader should be prepared for the main argument in X, 10, 13-16for the conclusion that the mind is an immaterial substance – whichis what Augustine said would be clear to anyone who wasn’t suffer-ing from conceptual poverty in X, 7, 10 (cited in n. 26). So now hetakes up the case of an open-minded materialist reader exemplifiedby Cicero in Tusculan Disputations, I – the sort of reader character-ized in X, 5, 7-7, 10 and exercised in X, 7, 10-9, 13. And he argues thatgiven what we have seen about the mind’s necessary presence to it-self (its unique mode of presentation), Cicero’s own main argumentin Tusc. disp., I is sufficient to show him that the essential propertiesof the mind disclosed in the process of self-inquiry actually rule outmaterialism about the mind.38

Augustine starts the main argument by picking up the case ofthe materialists again in referring back to the dispute betweenthem about the nature of the mind, which he had outlined in hisdoxography in De trinitate, X, 7, 9. He begins with three introduc-tory points in X, 10, 13-15. First (X, 10, 13), he sketches the situationof the materialists in the light of exercises we have just completed:they are quite aware of the mental properties that in fact consti-

37. Augustine may think that he has dealt with the first problem conclusively inX, 3, 5-4, 6, with his arguments that the mind knows what it knows « as a whole ».But it is notable that he tries again in a fairly direct way in trin., X, 11, 17-18 (cf. IX,3, 3-5, 8); and his later treatments of intellectual memory, notably at XII, 14, 22-15, 24and XIV, 5, 7-11, 14, should be seen as further indirect treatments of this problem.He sets out an explicit version of the second problem in trin., XV, 7, 11-8, 14 & 21,40-24, 44.

38. Note that Augustine’s argument is not an anti-sceptical one, pace e.g.Schmaus, Die psychologische Trinitätslehre, 243, and Ayres, The Discipline of Self-knowl-edge, 278-279. Rather, it depends on the materialist interlocutor’s willingness toagree that his or her materialist theory of mind is less certain than our phenome-nological evidence for self-reflexivity.


tute the immaterial nature of the mind, but their metaphysicaland conceptual poverty leaves them no option but to “refer” theseproperties to something material.39 Then (X, 10, 14), he points out,following Cicero’s main argument in Tusc. disp., I, that their in-quiry into the nature of the mind leaves them in a quandary: onthe one hand they have a range of material candidates for the na-ture of the mind, about which there is no agreement or certainty;and on the other, there is no doubt about the set of mental prop-erties Cicero had characterized as “divine”, viz. the properties pre-supposed by the act of inquiry itself.

But who doubts that he is alive, remembers, understands, wants, thinks,knows and judges? After all, even if he doubts, he is alive; and if hedoubts: he remembers the source of his doubting; he understands that hedoubts; he wants to be certain; he is thinking; he knows that he doesn’tknow; he judges that he shouldn’t assent rashly. So anyone who doubtsfor some reason should not be in doubt about any of these things, giventhat if they weren’t the case, he couldn’t doubt about anything.40

(Scholars have taken the supporting argument in this passage asevidence that Augustine held some special views about knowledge– the “KK” principle, e.g. or the “conscientia doctrine”.41 But if we read

39. Note that Augustine’s cryptic reference to “referring” mental propertiesmistakenly to material stuffs in X, 10, 13 (which has puzzled all the translators andcommentators I am aware of, including even E. Bermon, Le « cogito », 357-359) be-comes clear in the light of Cic., Tusc. disp., I, 16, 37-38, i.e. in the light of his error the-ory. This passage thus supports the view that Augustine’s argument here is direct-ed at or in dialogue with Cicero.

40.Aug., trin., X, 10, 1438-45 (CCL 50, 327-328): «Viuere se tamen et meminisse etintellegere et uelle et cogitare et scire et iudicare quis dubitet? Quandoquidem eti-am si dubitat, uiuit; si dubitat, unde dubitet meminit; si dubitat, dubitare se intelle-git; si dubitat, certus esse uult; si dubitat, cogitat; si dubitat, scit se nescire; si dubi-tat, iudicat non se temere consentire oportere. Quisquis igitur alicunde dubitat dehis omnibus dubitare non debet quae si non essent, de ulla re dubitare non posset ».

41. For the “conscientia doctrine”, see A. Lloyd, Nosce teipsum and conscientia,«Arch. Gesch. Philos. », 46.2 (1964), 188-200; for variants on the “KK principle”, seeG. Matthews, Thought’s Ego in Augustine and Descartes, Cornell University Press, Ithaca(NY) 1992, 29-41; Id., Augustine, Blackwell, Oxford, 2005, 34-42, and Id., Augustine’s FirstPerson Perspective, in P. Cary - J. Doody - K. Paffenroth (cur.), Augustine and Philosophy,Lexington Books, Lanham 2010, 41-60, at 45-50; cf. Brittain, Intellectual Self-Knowledge.


it with its Ciceronian context in mind, the passage seems unre-markable: the argument is just the familiar one from Tusc. disp., I

that an act of inquiry presupposes a set of mental properties, activitiesand states.) Finally (X, 10, 15), he points out that whichever form ofmaterialism they adopt, materialists must assume that this set ofmental activities or properties are qualifications of or properties in amaterial substrate, rather than constituting an immaterial substancein their own right (as Augustine said they did in X, 10, 13).

To this point, Augustine has just restated the premises of Ci-cero’s main argument, albeit in a way that highlights what hetakes to be a crucial tension in Cicero’s or any materialist’s posi-tion. But inDe trinitate, X, 10, 16 he uses these premises to build thetwo formal arguments that represent his version of the main argu-ment.42 The first, the substance argument, goes like this:

All of these materialists have failed to note that the mind knows itselfeven when it is seeking itself, as I have already shown. But there’s no waythat something can be properly said to be known while its substance isnot known. Accordingly, when the mind knows itself it knows its sub-stance, and when it is certain about itself it is certain about its substance.But it is certain about itself, as the points I’ve made above show. But it isnot in the least certain whether it is air or fire or a body or something be-longing to a body. So it isn’t any of these things.43

This can be schematized thus:

1. The mind always knows itself (see X, 1, 1-4, 6; cf. X, 5, 7-9, 13).

2. If you know x, you know the essence of x.44

42. For more detailed treatments of these arguments, see the exemplary work ofG. Matthews, Augustine on the mind’s search for itself, «Faith Philos. », 20.4 (2003), 415-429;Id., Augustine, 43-52, and Brachtendorf, Die Struktur des menschilichen Geistes, 181-186.

43. Aug., trin., X, 10, 1661-68 (CCL 50, 328): «Qui omnes non aduertunt mentemnosse se etiam cum quaerit se sicut iam ostendimus. Nullo modo autem recte dicitursciri aliqua res dum eius ignoratur substantia. Quapropter dum se mens nouit subs-tantiam suam nouit, et cum de se certa est de substantia sua certa est. Certa est autemde se sicut conuincunt ea quae supra dicta sunt. Nec omnino certa est utrum aer anignis sit an aliquod corpus uel aliquid corporis. Non est igitur aliquid eorum».

44. Note that Augustine says repeatedly in trin. that ‘substantia’ and ‘essentia’ aretwo terms for the same thing, i.e. the nature that in God’s case is triune. (On the


<3. If you know the essence of x, you are certain of the essentialproperties of x.>

4. So the mind knows its essence and is certain of its essentialproperties.

5. But the mind is not certain of its having any material properties(see X, 10, 13-14).

6. So the mind does not have any essential material properties, i.e.is not material.

Since premises [1] and [5] have already been established (andI have added [3] just to make the connection between the termsclear), the crucial new premise here is [2]: the principle that know-ing something implies knowing its essence. This is a very strongepistemic principle, and not one found in Cicero or argued for byAugustine in De trinitate. It is unclear why the materialist readershould accept it; and even if she does, it doesn’t seem very helpfulhere. For if we accept premise [2], we are very unlikely to acceptpremise [1] without equivocation, since the arguments for themind’s knowledge of itself don’t seem to rely on the same strongsense of knowledge. This is clear from Cicero’s original version ofthe argument, where its point was precisely that we are certain ofand know some of the mind’s essential properties, even though wedo not know its nature or essence as such. In Augustine’s version,he has shown us that the mind knows itself as a subject, but notthat this exhausts its nature. So Augustine’s first revision of Cicero’smain argument doesn’t seem to secure the conclusion he wants.45

meaning of ‘substantia’ in trin., see R. Teske, Augustine’s use of ‘Substantia’ in Speakingabout God, «Mod. Schoolman», 62 (1985), 147-163.) Matthews,Augustine, 45-46 suggeststhat we can read ‘substantia’ as meaning either “stuff” or “essence” here. The “stuff ’reading can find some basis in the Ciceronian context and in trin., X, 7, 9, but is not thesynonym of “nature” that Cicero and Augustine took themselves to be discussing.In any case, it doesn’t help here, since [2] is no more plausible if it means “stuff”.

45. We can salvage Augustine’s honour here by taking him to mean by premise[1] what he actually argued for in X, 4, 6, viz. that the mind always has complete self-knowledge. Premise [2] would then be just a poorly expressed point that if weknow something completely then we know its essence.


His second version, the mode argument, is more complicated:

For it thinks of fire just as it thinks of air and any other material thing, andthere’s no way that it could think what it is itself in the same way as itthinks what it is not itself. Because it thinks all these material things [...]through representations. But if it were one of them, it would think this ina different way than it thought the others, viz. not through a representa-tional state, in the way one thinks about absent things <through percep-tion or imagination> [...] Rather, it would think it owing to an inner pres-ence that wasn’t simulated, but a real presence – given that nothing ismore present to the mind than the mind itself. <So it would think it> inthe way that it thinks that it is alive and that it remembers and under-stands and wants. For it knows these things in itself and doesn’t representthem as if they were external things it has grasped by a sense in the waymaterial things are grasped. And if it stops adding to itself on the basis ofits thoughts about such things so that it thinks it is itself something of thissort, what is left in its thoughts about itself is just what it is.46

The basic argument might be regimented as follows:

7. The mind is entirely present to itself.

8. It is not present to itself as something material.

9. So the mind is not material.

This bare-bones version highlights the main problem Cicero– among other readers – would find in the argument, which is,again, the assumption that Augustine has shown us that all the es-sential properties of the mind are present to it or disclosed to it in

46.Aug., trin., X, 10, 1671-87 (CCL 50, 329): « Sic enim cogitat ignem ut aerem etquidquid aliud corporis cogitat, neque ullo modo fieri posset ut ita cogitaret idquod ipsa est quemadmodum cogitat id quod ipsa non est. Per phantasiam quippeimaginariam cogitat haec omnia, siue ignem siue aerem siue illud uel illud corpuspartemue ullam seu compaginem temperationemque corporis, nec utique ista om-nia sed aliquid horum esse dicitur. Si quid autem horum esset, aliter id quam ceteracogitaret, non scilicet per imaginale figmentum sicut cogitantur absentia quae sen-su corporis tacta sunt, siue omnino ipsa siue eiusdem generis aliqua, sed quadaminteriore non simulata sed uera praesentia (non enim quidquam illi est se ipsa prae-sentius), sicut cogitat uiuere se et meminisse et intellegere et uelle se. Nouit enimhaec in se, nec imaginatur quasi extra se illa sensu tetigerit sicut corporalia quaequetanguntur. Ex quorum cogitationibus si nihil sibi affingat ut tale aliquid esse seputet, quidquid ei de se remanet hoc solum ipsa est ».


self-reflection or just in conscious inquiry or thought. For it is notclear that he has given us good reasons to think that the mindmight not have additional properties that underlie and ground itsconsciousness – whether these turn out to be further mental dis-positions (as I suggested above) or, for all we know, materialgrounds for its conscious states.47

But Augustine’s introduction of the second passage in De trini-tate, X, 10, 1668-71 indicates that its purpose is not to give a secondreason to think that the soul is notmaterial, but rather to reach thepositive conclusion that the soul is immaterial because it is identicalto the set of essential properties it is certain of (the set Cicerocalled “divine”).48 The core idea in the second argument is thus,I think, the rather different one that the mind’s necessary presenceto itself makes its mode of self-presentation completely differentto the modes under which other and especially material things arepresented to it. Augustine puts this in two ways here. The first iscounterfactual: if it were material x, it would necessarily think ofits x-ness – itself – in a way unlike the way in which it thinks aboutanother material, e.g. y or about y-ness. The second is positive:the mind’s mode of self-presentation is an inner and real or un-simulated presence, which doesn’t involve any representation.This is what the three exercises in X, 7, 10-9, 13 were supposed todisclose to us, viz. that the mind is already “in” the mind, and right“inside”, and so immediately “present” to itself.

It’s not entirely clear if Augustine means us to take these twoways of explaining his basic insight here as making different

47. This is not to say that Augustine hasn’t succeeded in showing that the mindcan’t be purely material, which, if true, is an impressive result; see again the excel-lent reconstructions of his argument in Matthews Augustine, 50-52, & Augustine onthe mind’s search for itself, 424-428. I disagree with him only concerning his objectionthat we may have “unconscious” desires or beliefs in a Freudian sense: such desires,if we have them, would be manifested in the lower soul, not in the “mind” underdiscussion here.

48. Aug., trin., X, 10, 1668-71 (CCL 50, 328-329): «And the whole point of its beingtold to know itself is for it to be certain that it is not any of the things it is uncertainabout and for it only to be certain that it is only what it is certain about (Totumqueillud quod se iubetur ut nouerit, ad hoc pertinet ut certa sit non se esse aliquid eorum de quibusincerta est, idque solum esse se certa sit quod solum esse se certa est) ».


points. But it seems reasonable to suppose that the second way in-dicates that is a mistake to think that there are in general two kindsof awareness or two “modes of presentation”, as I have been put-ting it, as if the mind were just another sort of object for cognition.Rather, the right way to think about the mind’s self-knowledge isthat it is a general condition of its being the subject of thoughts inthe first place.49 If so, Augustine’s second version of Cicero’s mainargument is more successful in capturing something vital aboutself-inquiry that was perhaps implicit in Cicero’s original butneeded Augustine to bring it to the fore. This would be a worth-while result, I think, even if the argument as a whole seems des-tined to fail for the same reason as the first version: we really don’tknow what underlies conscious thought.

There is much more to say about these arguments. But I willconclude with a few words on Augustine’s use of the three Ci-ceronian elements we have been examining. In each case, what wefind, I think, is that Augustine’s reading of Cicero yields an inten-sification of the argument. In the case of the error theory, he takesa mild point about the Epicureans’ failure to conceive the exis-tence of the (material) soul when the body perishes, and translatesit into a critique of materialism about the mind, as a form of con-ceptual poverty that fails to make due allowance for the set of irre-ducibly immaterial properties that we all know characterize themind (whatever its essence may be). Likewise with the appeal tosubjectivity and the reader’s own case: Augustine picks up somehints in Cicero’s text and shapes them into the basis of his protrep-tic method. And again with the “main argument”: Cicero’s modestbut significant argument that the presuppositions of the mind’sself-inquiry show some of its essential properties becomes, at Au-gustine’s hands, an argument that discloses its nature entirely.

49.Note, though, that Augustine argues in trin., XIV, 6, 8, that it is possible, ifvery unusual, for the mind to take itself as an immediate (and sole) “object”, in acase of self-intellection, as I argue in Brittain, Intellectual Self-Knowledge, 334-339.


The last case may be a step too far. But perhaps it is enough forAugustine’s wider purpose in De trinitate to have shown his slowerreaders that they have direct knowledge of the sort of self-reflex-ive immaterial properties of the mind that he wants to use as theimage of the divine Trinity in Books XIV & XV. At any rate, it isdefinitely enough, I think, to show that Augustine was an excel-lent reader of Cicero.50

50. I am very grateful to the editors of this volume and to the audiences of earli-er versions of this paper in London and Bonn, and especially to Giovanni Cata-pano, Christoph Horn and Scott MacDonald. I am also delighted to acknowledgethe generosity of the School of Advanced Studies at London for their support of myresearch on Augustine through the S.T. Lee Professorial Fellowship for 2010-2011.


Self-knowledge in Cicero and Augustine (De Trinitate 10)

Documents