In the first part of the article, The art of memory, I exposed the theory of loci, perhaps the oldest complex mnemo-technique that we know. We have to say that loci – “places” where the images need to be placed – are well defined spaces. However, they are spaces that we must build only in our mind. They are images, and that of “images of memory” is a pivotal concept for us.
Since places and theatres are a special kind of image, we have to gather that the concept of “image” is the fundamental one in the art of memory. One of the most widespread beliefs credence about imagination is that the mental images are like photos. This is a serious mistake for a lot of reasons. The first one is that our imagination is less defined than a camera. The second one is that our imagination is a living thing since a camera and its photos are inanimate. The third reason is that our imagination doesn’t reproduce reality, but rebuilds every time that which the mind knows. This rebuilding work is not a replica of what happens “out there”, it isn’t actually a replica at all. Imagination is an incessant job of replacing; it’s a continuous adjustment and adaptation between our structured knowledge, on one side – and all the new things we know (and/or think at any time of our conscious life), on the other side.
The correct juxtaposition of the images of things, those of the places and that of the theatre is the base of the memory for things (memoria rerum), but mnemonists have devised other systems to memorize words (memoria verborum). The latter is less functional and more difficult to learn. For us, at the moment, they are non essential. I mention them only for some elements that can be useful to figure out the rest of this work. Sometimes mnemonists use letters for meaning concepts or things or other objects of thought. Sometimes they take advantage of similitude in the sounds of words, or in visualizing the shapes of things, etc. Other times they contrive mnemonic vocabularies based on objects whose names start with the same letter of the alphabet (for instance: albatross for ‘A’; Bald eagle for ‘B’; Canary for ‘C’ and so on). Anyway, there is always a kind of association between the letters (or the words) to remember and the things that they (the letters or the words) cross-reference in the imagination that can be changed from one mnemonic system to another. Then again, it’s the same thing that happens when children learn to read and write through the visual alphabets. Some advertisements use this kind of association (combined with other strategies) in order to give more impact to the messages. I show this sort of variable where it is appropriate.
Ultimately at the core of the debate there is always the concept of mental image. This is where we have to start to understand the theory of loci and, by extension, the art of memory as a whole. Indeed, also the theatre of memory and its places are special kinds of images. But, what kind of images are the mental images of memory? Obviously they are mental images like all the others. However, not all the mental images are suitable for a mnemonic use since there are particular images that stay in mind better than others. As regards the previous distinction between memoria rerum and memoria verborum F. A. Yates says: «“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 memorizing every single word in it in the right order? The first kind of artificial memory is memoria rerum; the second kind is memoria verborum» . I insist on this disjunction since it is at the base of all the uses of mental images. Usually we don’t use the same kind of images for both memoria rerum and memoria verborum: to achieve different purposes, we have to use different tools. However, there are a few features that all images must have whatever the nature of memories is. Although when we refer to the term imagination, we think immediately of the visual and the figurative one, we must take due account that there is also an olfactory, a tactile imagination, etc. There can be no denying, nonetheless, that the visual imagination is the most important one. This is what enables a more rapid and unconscious abstraction from the image to the concept. I can abstract the concept of a circle on the shape of the cheese, but not by its smell if not “by proxy of another image”; that is, by the smell of cheese I move to visualize the shape and from this latter I move toward the representation of the circle and from here to the concept of the composed geometric circle. To a first approximation we can say that visual imagination produces complex images: from the art of memory point of view, it consists in a foreground and a background, color management and the definition of the forms. A (mnemonic) mental image is never a faithful reproduction of what is perceived. The more it is rich in detail, the more it resembles a perceived thing; but a mnemonic image and a perception are never the same. Instead, when a mental image is lacking in detail, and we consider it only for its shape and for a few other distinguishing features, it means that abstraction plays an important cognitive function. In this case, the image becomes more and more different from the perception, and the more it moves away from the thing experienced, the more it approaches the concept of the thing. Usually, the images of the concrete things are easier to remember than abstract ones, as we saw in a previous article . Anyway, this is not a fixed rule, for example: there are many symbols that are abstract “stylizations” of concrete images and if they are very well assimilated (as in the cases of some religious, political symbols and so on) they can easily be retrieved when necessary. However, in these cases there is always an iconic tie that facilitates the semantic memorization of the abstract thing. A different case is when there isn’t a concrete link between the image, the symbol and the thing that the symbol denotes. This is not always a problem. In the case of mathematical thinking it is indeed a very positive feature and this prevents the mind from being filled with images which are not only unnecessary, but which even slow down the development of the thoughts and the argumentation.
Now I will try to develop a reasoning that seems to contradict what I've written so far, but it is not exactly so. We certainly have to imagine things in the clearest and most distinct manner possible, but since it is impossible and not desirable to perfectly reproduce the thing in itself, the images should be completed only for some few features that facilitate the memorization. For example: if you have to remember to take a jacket to the laundry, it is not necessary to imagine the washerwoman perfectly. You have to imagine her only in her essential traits . However, these latter must be thought precisely, then our mind reconstructs the overall image independently. If I have to remind myself to get gas for my car, I can imagine the attendant of the gas station where I usually go. I will have to think about that specific person, but I should not depict him in any specific detail: the cap, the jacket and some of his distinctive facial features will be sufficient to fix the memory (get gas for the car) in my mind. In this way I can build a memory without the mediation of language. Implicitly I memorize a sentence («I have to fill up with fuel»), but using only images and not words. So, I can avoid making an effort of abstraction, namely an effort to memorize.
There are certain characteristics of the images that serve better than others to remember the images themselves in their entirety. To move from thinking of the cheese in its geometric shape, we need to make abstraction from a concrete image (cheese) to an abstract representation of a geometrical figure (the circle). We have to take something away from the image. This implies a high intellectual effort that is tiring. On the contrary, when we think of a concrete thing starting from some of its features, our mind naturally tends to complete the overall image. In this case, our mind has to add something. If we have chosen well to highlight the details, this task is not very tiring; indeed sometimes it does not require any effort. But we must be careful not to make a mistake: we have to avoid thinking only of details and by doing so, replacing the thing itself with its characteristics. Namely, I don’t have to think of the cap and the work jacket of Mario (the attendant of the gas station), but of Mario himself with his hat and his jacket while he fills up my car. It is not questioned to replace something with something else, but rather focus on something to rebuild the whole picture. The incompleteness of the image must be weighed; it must not be something which is missing, but rather an approximation. The mnemonic representation must clearly indicate the object, but without reproducing it perfectly. It must accurately indicate, but only indicate, letting it be our mind to rebuild the complete memory. The mnemonic image should look like a first sketch of a Renaissance painter rather than a hyper-realistic portrait executed on the model of the Flemish Baroque painting school. The sketch gives the idea of the thing that we have to memorize without weighing down the memory.
 F.A. Yates, The art of memory, Routledge & Kogan Paul, London - New York, 1966 p. 9.
 Memory and the “dual coding” theory
 I mean essential for the information that we need to remember. It is irrelevant visualize the washerwoman color of the eyes or the shape of the ears. It's better if I imagine her with my jacket in his left hand and the electric iron in his right hand.