THE ART OF MEMORY

One of Modern Library's 100 Best Nonfiction Books of the Twentieth Century

In this classic study of how people learned to retain vast stores of knowledge before the invention of the printed page, Frances A. Yates traces the art of memory from its treatment by Greek orators, through its Gothic transformations in the Middle Ages, to the occult forms it took in the Renaissance, and finally to its use in the seventeenth century. This book, the first to relate the art of memory to the history of culture as a whole, was revolutionary when it first appeared and continues to mesmerize readers with its lucid and revelatory insights.

The Art of Memory according to Ad Herennium

Frances Yates

An unknown teacher of rhetoric in rome complied, circa 86-82 B.C., a useful text-book for his students which immortalised, not his own name, but the name of the man to whom it was dedicated. It is somewhat tiresome that this work, so vitally important for the history of the classical art of memory [...] has no other title save the uniformative Ad Herennium. The busy and efficient teacher goes through the five parts of rhetoric (inventio, dispositio, elocutio, memoria, pronuntiatio) in a rather dry text-book style. When he comes to memory as an essential part of the orator's equipment, he opens his treatment of it with the words: 'Now let us turn to the treasure-house of inventions, the custodian of all the parts of rhetoric, memory.' There are two kinds of memory [Ad Herennium claims], one natural, the other artificial. The natural memory is that which is engrafted in our minds, born simultaneously with thought. The artificial memory is a memory strengthened or confirmed by training. A good natural memory can be improved by this discipline and persons less well endowed can have their weak memories improved by the art.

After this curt preamble the author announces abruptly, 'Now we will speak of the artificial memory.'

An immense weight of history presses on the memory section of Ad Herennium. It is drawing on Greek sources of memory teaching, probably in Greek treatises on rhetoric all of which are lost. It is the only Latin treatise on the subject to be preserved, for Cicero's and Quintilian's remarks are not full treatises and assume that the reader is already familiar with the artificial memory and its terminology. It is thus really the main source, and indeed the only complete source, for the classical art of memory both in the Greek and in the Latin world. Its rôle as the transmitter of the classical art to the Middle Ages and the Renaissance is also of unique importance. The Ad Herennium was a well known and much used text in the Middle Ages when it had an immense prestige because it was thought to be by Cicero. It was therefore believed that the precepts for the artificial memory which it expounded had been drawn up by 'Tullius' himself.

In short, all attempts to puzzle out what the classical art of memory was like must be mainly based on the memory section of Ad Herennium. And all attempts such as we are making in this book to puzzle out the history of that art in the Western tradition must refer back constantly to this text as the main source of the tradition. Every Ars memorativa treatise, with its rules for 'places', its rules for 'images', its discussion of 'memory for things' and 'memory for words', is repeating the plan, the subject matter, and as often as not the actual words of Ad Herennium. And the astonishing developments of the art of memory in the sixteenth century [...] still preserve the 'Ad Herennian' outlines below all their complex accretions. Even the wildest flights of fancy in such a work as Giordano Bruno's I cannot conceal the fact that the philosopher of the Renaissance is going through yet once again the old, old business of rules for places, rules for images, memory for things, memory for words.

Evidently, therefore, it is incumbent upon us to attempt the by no means easy task of trying to understand the memory section of Ad Herennium. What makes the task by no means easy is that the rhetoric teacher is not addressing us; he is not setting out to explain to people who know nothing about it what the artificial memory was. He is addressing his rhetoric students as they congregated around him circa 86-82 B.C., and they knew what he was talking about; for them he needed only to rattle off the 'rules' which they would know how to apply. We are in a different case and are often somewhat baffled by the strangeness of some of the memory rules.
In what follows I attempt to give the content of the memory section of Ad Herennium, emulating the brisk style of the author, but with pauses for reflection about what he is telling us.

The artificial memory is established from places and images (Constat igitur artificiosa memoria ex locis et imaginibus), the definition to be forever repeated down the ages. A locus is a place easily grasped by the memory, such as a house, an intercolumnar space, a corner, an arch, or the like. Images are forms, marks or simulacra (formae, notae, simulacra) of what we wish to remember. For instance if we wish to recall the genus of a horse, of a lion, of an eagle, we must place their images on definite loci.

The art of memory is like an inner writing. Those who know the letters of the alphabet can write down what is dictated to them and read out what they have written. Likewise those who have learned mnemonics can set in places what they have heard and deliver it from memory. 'For the places are very much like wax tablets or papyrus, the images like the letters, the arrangement and disposition of the images like the script, and the delivery is like the reading.'

If we wish to remember much material we must equip ourselves with a large number of places. It is essential that the places should form a series and must be remembered in their order, so that we can start from any locus in the series and move either backwards or forwards from it. If we should see a number of our acquaintances standing in a row, it would not make any difference to us whether we should tell their names beginning with the person standing at the head of the line or at the foot or in the middle. So with memory loci. 'If these have been arranged in order, the result will be that, reminded by the images, we can repeat orally what we have committed to the loci, proceeding in either direction from any locus we please.'

The formation of the loci is of the greatest importance, for the same set of loci can be used again and again for remembering different material. The images which we have placed on them for remembering one set of things fade and are effaced when we make no further use of them. But the loci remain in the memory and can be used again by placing another set of images for another set of material. The loci are like the wax tablets which remain when what is written on them has been effaced and are ready to be written on again.

In order to make sure that we do not err in remembering the order of the loci it is useful to give each fifth locus some distinguishing mark. We may for example mark the fifth locus with a golden hand, and place in the tenth the image of some acquaintance whose name is Decimus. We can then go on to station other marks on each succeeding fifth locus.

It is better to form one's memory loci in a deserted and solitary place for crowds of passing people tend to weaken the impressions. Therefore the student intent on acquiring a sharp and well-defined set of loci will choose an unfrequented building in which to memorise places.

Memory loci should not be too much like one another, for instance too many intercolumnar spaces are not good, for their resemblance to one another will be confusing. They should be of moderate size, not too large for this renders the images placed on them vague, and not too small for then an arrangement of images will be overcrowded. They must not be too brightly lighted for then the images placed on them will glitter and dazzle; nor must they be too dark or the shadows will obscure the images. The intervals between the loci should be of moderate extent, perhaps about thirty feet, 'for like the external eye, so the inner eye of thought is less powerful when you have moved the object of sight too near or too far away'.

A person with a relatively large experience can easily equip himself with as many suitable loci as he pleases, and even a person who thinks that he does not possess enough sufficiently good loci can remedy this. 'For thought can embrace any region whatsoever and in it and at will construct the setting of some locus.' (That is to say, mnemonics can use what were afterwards called 'fictitious places', in contrast to the 'real places' of the ordinary method.)

Pausing for reflection at the end of rules for places I would say that what strikes me most about them is the astonishing visual precision which they imply. In a classically trained memory the space between the loci can be measured, the lighting of the loci is allowed for. And the rules summon up a vision of a forgotten social habit. Who is that man moving slowly in the lonely building, stopping at intervals with an intent face ? He is a rhetoric student fomiing a set of memory loci.

'Enough has been said of places', continues the author of Ad Herennium, 'now we turn to the theory of images.' Rules for images now begin, the first of which is that there are two kinds of images, one for 'things' (res), the other for 'words' (verba). That is to say 'memory for things' makes images to remind of an argument, a notion, or a 'thing'; but 'memory for words' has to find images to remind of every single word.

I interrrupt the concise author here for a moment in order to remind the reader that for the rhetoric student 'things' and 'words' would have an absolutely precise meaning in relation to the five parts of the rhetoric. Those five parts are defined by Cicero as follows:

Invention is the excogitation of true things (res), or things similar to truth to render one's cause plausible; disposition is the arrangement in order of the things thus discovered; elocution is the accomodation of suitable words to the invented (things); memory is the firm perception in the soul of things and words; pronunciation is the moderating of the voice and body to suit the dignity of the things and words. [De inventione, I, vii, 9]

'Things' are thus the subject matter of the speech; 'words' are the language in which that subject matter is clothed. Are you aiming at an artificial memory to remind you only of the order of the notions, arguments, 'things' of your speech ? Or do you aim at memorising every single word in it in the right order ? The first kind of artificial memory is memoria rerum; the second kind is memoria verborum. The ideal, as defined by Cicero in the above passage, would be to have a 'firm perception in the soul' of both things and words. But 'memory for words' is much harder than 'memory for things'; the weaker brethren among the author of Ad Herennium's rhetoric students evidently rather jibbed at memorising an image for every single word, and even Cicero himself, as we shall see later, allowed that 'memory for things' was enough.

To return to the rules for images. We have already been given the rules for places, what kind of places to choose for memorising. What are the rules about what kind of images to choose for memorising on the places? We now come to one of the most curious and surprising passages in the treatise, namely the psychological reasons which the author gives for the choice of mnemonic images. Why is it, he asks, that some images are so strong and sharp and so suitable for awakening memory, whilst others are so weak and feeble that they hardly stimulate memory at all? We must enquire into this so as to know which images to avoid and which to seek.

Now nature herself teaches us what we should do. When we see in every day life things that are petty, ordinary, and banal, we generally fail to remember them, because the mind is not being stirred by anything novel or marvellous. But if we see or hear something exceptionally base, dishonourable, unusual, great, unbelievable, or ridiculous, that we are likely to remember for a long time. Accordingly, things immediate to our eye or ear we commonly forget; incidents of our childhood we often remember best. Nor could this be so for any other reason than that ordinary things easily slip from the memory while the striking and the novel stay longer in the mind. A sunrise, the sun's course, a sunset are marvellous to no one because they occur daily. But solar eclipses are a source of wonder because they occur seldom, and indeed are more marvellous than lunar eclipses, because these are more frequent. Thus nature shows that she is not aroused by the common ordinary event, but is moved by a new or striking occurrence. Let art, then, imitate nature, find what she desires, and follow as she directs. For in invention nature is never last, education never first; rather the beginnings of things arise from natural talent, and the ends are reached by discipline.

We ought, then, to set up images of a kind that can adhere longest in memory. And we shall do so if we establish similitudes as striking as possible; if we set up images that are not many or vague but active (imagines agentes); if we assign to them exceptional beauty or singular ugliness; if we ornament some of them, as with crowns or purple cloaks, so that the similitude may be more distinct to us; or if we somehow disfigure them, as by introducing one stained with blood or soiled with mud or smeared with red paint, so that its form is more striking, or by assigning certain comic effects to our images, for that, too, will ensure our remembering them more readily. The things we easily remember when they are real we likewise remember without difficulty when they are figments. But this will be essential - again and again to run over rapidly in the mind all the original places in order to refresh the images. [Ad Herennium, III, xxii]

Our author has clearly got hold of the idea of helping memory by arousing emotional affects through these striking and unusual images, beautiful or hideous, comic or obscene. And it is clear that he is thinking of human images, of human figures wearing crowns or purple cloaks, bloodstained or smeared with paint, of human figures dramatically engaged in some activity - doing something. We feel that we have moved into an extraordinary world as we run over his places with the rhetoric student, imagining on the places such very peculiar images. Quintilian's anchor and weapon as memory images, though much less exciting, are easier to understand than the weirdly populated memory to which the author of Ad Herennium introduces us.

It is one of the many difficulties which confront the student of the history of the art of memory that an Ars memorativa treatise though it will always give the rules, rarely gives any concrete application of the rules, that is to say it rarely sets out a system of mnemonic images on their places. This tradition was started by the author of Ad Hennnium himself who says that the duty of an instructor in mnemonics is to teach the method of making images, give a few examples, and then encourage the student to form his own. When teaching 'introductions', he says, one does not draft a thousand set introductions and give them to the student to learn by heart; one teaches him the method and then leaves him to his own inventiveness. So also one should do in teaching mnemonic images. This is an admirable tutorial principle though one regrets that it prevents the author from showing us a whole set or gallery of striking and unusual imagines agentes. We must be content with the three specimens which he describes.

The first is an example of a 'memory for things' image. We have to suppose that we are the counsel for the defence in a law suit. 'The prosecutor has said that the defendant killed a man by poison, has charged that the motive of the crime was to gain an inheritance, and declared that there are many witnesses and accessories to this act.' We are forming a memory system about the whole case and we shall wish to put in our first memory locus an image to remind us of the accusation against our client. This is the image.

We shall imagine the man in question as lying illin bed, if we know him personally. If we do not know him, we shall yet take some one to be our invalid, but not a man of the lowest class, so that he may come to mind at once. And we shall place the defendant at the bedside, holding in his right hand a cup, in his left, tablets, and on the fourth finger, a ram's testicles. In this way we can have in memory the man who was poisoned, the witnesses, and the inheritance.

The cup would remind of the poisoning, the tablets, of the will or the inheritance, and the testicles of the ram through verbal similarity with testes - of the witnesses. The sick man is to be like the man himself, or like someone else whom we know (though not one of the anonymous lower classes). In the following loci we would put other counts in the charge, or the details of the rest of the case, and if we have properly imprinted the places and images we shall easily be able to remember any point that we wish to recall.

This, then, is an example of a classical memory image - consisting of human figures, active, dramatic, striking, with accessories to remind of the whole 'thing' which is being recorded in memory. Though everything appears to be explained, I yet find this image baffling. Like much else in Ad Herennium on memory it seems to belong to a world which is either impossible for us to understand or which is not being really fully explained to us.

The writer is not concerned in this example with remembering the speeches in the case but with recording the details or 'things' of the case. It is as though, as a lawyer, he is forming a filing cabinet in memory of his cases. The image given is put as a label on the first place of the memory file on which the records about the man accused of poisoning are kept. He wants to look up something about that case; he turns to the composite image in which it is recorded, and behind that image on the following places he finds the rest of the case. If this is at all a correct interpretation, the artificial memory would now be being used, not only to memorise speeches, but to hold in memory a mass of material which can be looked up at will.

The words of Cicero in the De oratore when he is speaking of the advantages of the artificial memory may tend to confirm this interpretation. He has just been saying that the loci preserve the order of the facts, and the images designate the facts themselves, and we employ the places and images like a wax writing tablet and the letters written on it. 'But what business is it of mine', he continues, 'to specify the value to a speaker and the usefulness and effectiveness of memory? of retaining the information given you when you were briefed and the opinions you yourself have formed? of having all your ideas firmly planted in your mind and all your resources of vocabulary neatly arranged, of giving such close attention to the instructions of your client and to the speech of the opponent you have to answer that they may seem not just to pour what they say into your ears but to imprint it on your mind? Consequently only people with a powerful memory know what they are going to say and for how long they are going to speak and in what style, what points they have already answered and what still remains; and they can also remember from other cases many arguments which they have previously advanced and many which they have heard from other people.'

We are in the presence of amazing powers of memory. And, according to Cicero, these natural powers were indeed aided by training of the type described in Ad Herennium.

The specimen image just described was a 'memory for things' image; it was designed to recall the 'things' or facts of the case and the following loci of the system would presumably have held other 'memory for things' images, recording other facts about the case or arguments used in speeches by the defence or the prosecution. The other two specimen images given in Ad Herennium are 'memory for words' images.

The student wishing to acquire 'memory for words' begins in the same way as the 'memory for things' student; that is to say he memorises places which are to hold his images. But he is confronted with a harder task for far more places will be needed to memorise all the words of a speech than would be needed for its notions. The specimen images for 'memory for words' are of the same type as the 'memory for things' image, that is to say they represent human figures of a striking and unusual character and in striking dramatic situations - imagines agentes.

We are setting out to memorise this line of verse:

Iam domum itionem reges Atridae parant
(And now their homecoming the kings, the sons of Atreus are making ready)

The line is found only in the quotation of it in Ad Herennium and was either invented by the author to exhibit his mnemonic technique or was taken for some lost work. It is to be memorised through two very extraordinary images.

One is 'Domitius raising his hands to heaven while he is lashed by the Marcii Reges'. The translator and editor of the text in the Loeb edition (H. Caplan) explains in a note that 'Rex was the name of one of the most distinguished families of the Marcian gens; the Domitian, of plebeian origin, was likewise a celebrated gens'. The image may reflect some street scene in which Domitius of the plebeian gens (perhaps bloodstained to make him more memorable) is being beaten up by some members of the distinguished Rex family. It was perhaps a scene which the author himself had witnessed. Or perhaps it was a scene in some play. It was a striking scene in every sense of the word and therefore suitable as a mnemonic image. It was put on a place for remembering this line. The vivid image immediately brought to mind 'Domitius-Reges' and this reminded by sound resemblance of 'domum itionem reges'. It thus exhibits the principles of a 'memory for words' image which brings to mind the words which the memory is seeking through their sound resemblance to the notion suggested by the image.

We all know how, when groping in memory for a word or a name, some quite absurd and random association, something which has 'stuck' in the memory, will help us to dredge it up. The classical art is systematising that process.

The other image for memorising the rest of the line is 'Aesopus and Cimber being dressed for the roles of Agamemnon and Menelaus in Iphigenaia'. Aesopus was a well-known tragic actor, a friend of Cicero; Cimber, evidently also an actor, is only mentioned in this text. The play in which they are preparing to act also does not exist. In the image these actors are being dressed to play the parts of the sons of Atreus (Agamemnon and Menelaus). It is an exciting off-stage glimpse of two famous actors being made up (to smear an image with red paint makes it memorable according to the rules) and dressed for their parts. Such a scene has an the elements of a good mnemonic image; we therefore use it to remember 'Atridae parant', the sons of Atreus are making ready. This image immediately gave the word 'Atridae' (though not by sound resemblance) and also suggested 'making ready' for the home-coming through the actors making ready for the
stage.

This method for memorising the verse will not work by itself, says the author of Ad Herennium. We must go over the verse three or four times, that is learn it by heart in the usual way, and then represent the words by means of images. 'In this way art will supplement nature. For neither by itself will be strong enough, though we must note that theory and technique are much the more rehable.' The fact that we have to learn the poem by heart as well, makes 'memory for words' a little less baffling.

Reflecting on the 'memory for words' images, we note that our author seems now concerned not with the rhetoric students' proper business of remembering a speech, but with memorising verse in poems or plays. To remember a whole poem or a whole play in this way one has to envisage 'places' extending one might almost say for miles within the memory, 'places' past which one moves in reciting, drawing from them the mnemonic cues. And perhaps that word 'cue' does give a clue to how the method might be workable. Did one really learn the poem by heart but set up some places with 'cue' images on them at strategic intervals ?

Our author mentions that another type of 'memory for words' symbol has been elaborated by the Greeks. 'I know that most of the Greeks who have written on the memory have taken the course of listing images that correspond to a great many words, so that persons who wished to learn these images by heart would have them ready without expending effort in a search for them.' It is possible that these Greek images for words are shorthand symbols or notae the use of which was coming into fashion in the Latin world at this time. As used in mnemonics, this would presumably mean that, by a kind of inner stenography, the shorthand symbols were written down inwardly and memorised on the memory places. Fortunately our author disapproves of this method, since even a thousand of such ready-made symbols would not begin to cover all the words used. Indeed, he is rather lenient about 'memory for words' of any kind; it must be tackled just because it is more difficult than 'memory for things'. It is to be used as an exercise to strengthen 'that other kind of memory, the memory for things, which is of practical use. Thus we may without effort pass from this difficult training to ease in that other memory.'

The memory section closes with an exhortation to hard work. 'In every discipline artistic theory is of little avail without unremitting exercise, but especially in mnemonics, theory is almost valueless unless made good by industry, devotion, toil, and care. You can make sure that you have as many places as possible and that these conform as much as possible to the rules; in placing the images you should exercise every day.'

We have been trying to understand inner gymnastics, invisible labours of concentration which are to us most strange, though the rules and examples of Ad Herennium give mysterious glimpses into the powers and organisation of antique memories. We think of memory feats which are recorded of the ancients, of how the elder Seneca, a teacher of rhetoric, could repeat two thousand names in the order in which they had been given; and when a class of two hundred students or more spoke each in turn a line of poetry, he could recite all the lines in reverse order, beginning from the last one said and going right back to the first. Or we remember that Augustine, also trained as a teacher of rhetoric, tells of a friend called Simplicius who could recite Virgil backwards. We have learned from our text-book that if we have properly and firmly fixed our memory places we can move along them in either direction, backwards or forwards. The artificial memory may explain the awe inspiring ability to recite backwards of the elder Seneca and of Augustine's friend. Pointless though such feats may seem to us, they illustrate the respect accorded in antiquity to the man with the trained memory.

Very singular is the art of this invisible art of memory. It reflects ancient architecture but in an unclassical spirit, concentrating its choice on irregular places and avoiding symmetrical orders. It is full of human imagery of a very personal kind; we mark the tenth place with a face like that of our friend Decimus; we see a number of our acquaintances standing in a row; we visualise a sick man like the man himself, or if we did not know him, like someone we do know. These human figures are active and dramatic, strikingly beautiful or grotesque. They remind one more of figures in some Gothic cathedral than of classical art proper. They appear to be completely amoral, their function being solely to give an emotional impetus to memory by their personal idiosyncracy or their strangeness. This impression may, however, be due to the fact that we have not been given a specimen image of how to remember, for example, the 'things' justice or temperance and their parts, which are treated by the author of Ad Herennium when discussing the invention of the subject matter of a speech. The elusiveness of the art of memory is very trying to its historian.