Alberto Cevolini

6.1 Openness and Closure in Secondary Memories

One of the most promising hypotheses of cultural studies is that media shape knowledge management. Consequently, media also shape how we think. Media represent a technical constraint on cognitive activity. Our understanding of this constraint is still evolving and unclear. As we work towards a general theory of media, it can be useful to investigate some historical reactions to the introduction of new media. In this essay, I focus on several effects that the printing press had on early modern Europe. Over the last fifty years, historical research has convincingly demonstrated that the advent of typographic technology compelled European scholars to drastically change their intellectual habits. During the period between 1550 and 1750, this change led to important educational developments and reflections regarding the method of studies and the methodological rules of a scholarly work. In turn, the printing press ensured that a subject whose emergence it had fostered would enjoy wide circulation.

One topic that aroused particular interest during this period was how to best read a book. The fact that this very question was the subject of a book and that, while reading, the reader was involved in a self-observation, however, went unnoticed. Scholars rediscovered a pillar of ancient and medieval education: the art of note-taking. This skill was deeply embedded in the rhetorical culture; consequently, the evolutionary advance of which the art of making excerpts from readings was simultaneously cause and effect was somehow misunderstood. Today, it is difficult to grasp the social nature of the long-lasting transitional stage that dismantled the old habits of European scholars and introduced the modern rules of studies. The contemporary sources must be read carefully to determine what makes a difference in meaning. If the social scientist is not to be deceived, he must discover where the contemporary authors deceived themselves, so to speak.

Before I address the sources upon which this study is based, it is helpful to quote an almost unknown text of a well-known learned man. This text offers a reflection that is missing in the prominent literature on the art of note-taking.

Its relevance might better be grasped from the perspective of the theory of modern society. In Lettre sur la Métode [sic] des Extraits, which was written during the first half of the eighteenth century, Charles-Irénée Castel, abbé de Saint Pierre praises the notes that his addressee (a noblewoman) has taken from moral books. The letter provides an opportunity for more general speculation on the (almost moral) usefulness of note-taking. Saint Pierre distinguishes between pleasure and usefulness and links this distinction to time. Pleasure is enjoyed in the present, whereas usefulness refers to the future. Notably, children are by nature prone to enjoy current pleasures because their reason is not yet ripe. By contrast, adults can appreciate future utility and are thus far wiser than children.1 Learned women who seek not only pleasurable but also useful thoughts and who wish to develop them in conversation with enlightened men should follow this advice. The practical device best suited to achieve this type of politesse is a commonplace-book.

The temporal habit praised by Saint Pierre is striking when we consider that in European culture until the seventeenth century, stoicism preserved its privileged position with respect to how time should be interpreted. From the perspective of this philosophical attitude, Saint Pierre reverses the order of priorities. According to stoicism, only the present belongs to man. Those who delay lose everything because they alienate the only good that men possess for themselves: the present. “Dum differtur vita transcurrit” (while they defer, life goes on), Seneca warned. True virtue is not to need future and is to avoid being preoccupied by worries about time that has yet to come.2

This statement is also true for memory, a faculty of the soul often associated with ethics until the end of the Middle Ages. In addition, memory is an internal good that should not be entrusted to an external support, according to Plato. To trust retrieval for when the need occurs is risky. One can be in danger – I return to this topic below – of lacking subject matters and arguments if suddenly the written support is not close at hand or if its hypomnematic function is disrupted.

When usefulness (i.e., the future) is valued more than pleasure (i.e., the present), a new temporal structure arises: the principle of deferred gratification,

which represents a pillar of modern society.3 American sociology of the 1950s linked this temporal structure to social stratification and regarded it as characteristic behaviour of the middle class. According to this approach, a middle-class citizen prefers to defer employment and economic independence in favour of earning a degree and thus more attractive employment opportunities in the future. In a sense, it is a question of relinquishing certain possibilities to create additional possibilities, increasing the complexity of the situation. In such abstract terms, the principle of postponing gratification is not only a facet of modern society but also an evolutionary principle. The most impressive examples involve money and law court (in place of self-satisfaction of punishment).

In the case of memory, the question is whether to relinquish the saturation of cognitive energies that is required to remember something, and to delay the construction of recollections. The weight of this relinquishment is better appreciated if one reminds that in the art of recollection the construction of a storehouse (thesaurus) – the architecture of the spaces in which the orator stored images that were to be used as mnemonic hooks – was a difficult task, but also a way of fixing memorable matters in the mind and keeping them at one’s immediate bidding. By contrast, by keeping a commonplace-book, early modern scholars committed themselves to performing a combinatory activity that could not be foreseen or pre-arranged (like the future itself) and that depended on opportunity. In addition, what was entrusted to the book was typically forgotten.

An advantage of this form of forgetfulness (that made it bearable) was that it released previously saturated cognitive energies and maintained them for subsequent use.4 In other words, what caused scholars to accept the disadvantages of relinquishing fixed notions in the mind of matters that were worth remembering was the seemingly paradoxical fact that relinquishing itself was advantageous. Social science should explain the circumstances under which such an improbable change became possible.

Implementing Saint Pierre’s advice in practice requires a memory that can cope with an open future. The card index seems to be a fitting solution. As with every secondary memory (e.g., archive or library), the card index is open in a twofold sense. First, it is understood as infinitely expandable – the expansion?

being performed either by adding new entries or by adding new items to existing entries. This limitlessness clearly distinguishes a filing cabinet from a rhetorical storehouse and was eagerly emphasized by early modern creators of filing systems. For instance, Jeremias Drexel was very proud of his note-taking method because it enabled students to daily add new headings. Thus, the number of entries (i.e., lemmata) in the commonplace-book could be endlessly increased (“in infinitum augeri possunt”).5 A half century later, Vincent Placcius stated that the advantage of excerpts in the form of loose file cards (schedacea excerpta) compared with bound commonplace-books was that a scholar could add new entries to his filing cabinet (“novis accessionibus semper augere possit”) wherever he was.6 In this sense, secondary memory is a historical machine because its content is contingent on the (not necessarily planned) sequence of readings and observations of its trainer. Moreover, its enlargement is in the first place physical: additional drawers are added to the filing cabinet and additional shelves or containers to the library.

However, the card index is also open in a different sense. The user may enlarge his collection of extracts, but he can also multiply the number of cross-references, links, and pointers (remissiones). Cross-references solve the problem of multiple storage, thus avoiding wasting time in copying an excerpt that must be stored under different entries because it has multiple meanings. References can also be associative; their function is to provide the card index with a self-referential closure and to structure new and surprising relationships.

Historical research has demonstrated that scholars became aware of this structural feature between the seventeenth and eighteenth centuries. For example, according to Christoph Meiners, the connection of facts and thoughts which are entrusted to a private card file can produce a substantial number of combinations and insights that otherwise might not have existed.7 Similarly, to legitimize his encyclopaedic work, Ephraim Chambers admitted that dismembering knowledge into loose entries compelled scholars to abandon their systematized arrangements, and one advantage of such dismembering was that?

scholars might occasionally discover relationships that they would otherwise not have noticed.8

The combinatory possibilities are directly proportional to the degree of loosening. Epistemologically speaking, this loosening implies the dissolution of the metaphysical order of knowledge. Practically speaking, this loosening coincides with bound notebooks being replaced by filing cabinets or scholarly chests. Early modern scholars understood that a secondary memory of this type might foster “Combinations and Coordinations of propositions” and that this manner of prompting recollections was in fact “the Argumentative part” of the card index. This memory practice eventually defused the primacy of syllogistic reasoning, that is, that form of rationality on which the art of recollection had been traditionally based.9

Despite certain hesitations and reasonable second thoughts, the cognitive saltation implied (and enabled) by card files was irreversible. The saltation was thematized at the end of a long-lasting transitional stage that developed until the second half of the seventeenth century. Robert Boyle noted that the arrangement of knowledge can be “either systematical, or more loose and unconfin’d”. The former solution implies the production of methodical treatises, whose real advantage is that they help memory. Their disadvantage is that science will quickly outgrow such treatises, as clothes suddenly no longer fit a child. The latter method of managing knowledge is of substantial value because it helps understanding. However, it can “scarce avoid the [sic] being plundered by systematical writers”, who can easily “cull out those things, that they like best, and insert them in their methodical books …”.10 Boyle did not?

look down on the efforts of compilers who produced “systems of a taking order”, yet he believed that the true scientist “discover(s) latent truths”, which may be best achieved when scholars relieve themselves – as Boyle did – of the burden of repetition and adopt a loose collection of observations that can be re-combined at will.

One of the most – if not the most – important cognitive resources is attention. In the art of reminiscence, attention was used to remember. Scholars made use of strange, odd, vivid ‘a(chǎn)ctive images’, according to the rule that familiar, trivial events do not elicit attention and are thus soon forgotten. Images were cues that triggered meaning associations that could presentify the past when correctly used – such as milk for white, white for air, air for wetness, and wetness for autumn in Aristotle’s famous example or as testicles (testes) in the hand in order to recall that there are witnesses (testes) of a murder, in the wellknown example of the rhetorical treatise to Herennius.11

If attention is relieved of this burden, it can be employed for different purposes. According to Descartes, who clearly understood this change, the advantage of entrusting all the certain cognitions that he grasped – from the simplest towards the more complicated – to a booklet instead of to his mind was that “with relieved memory I can turn my loosened mind to something else”.12 In a sense, the energies that are spared in this manner are conceived of as a form of intelligence that scholars in the seventeenth century first realized that they possessed. Such intelligence enabled them to discover remote neighbourhoods that had remained unrecognized or to notice inconsistencies that had been previously disregarded. The eventual oddity is that from this point forward, scholars paid attention to be surprised. Their opinion was: comparison is better than repetition, and what is unknown is more exciting than what is known. In short, from this point forward, memory was paradoxically employed to produce novelties.13

To achieve this result, memory must be stocked not only with contents but also with cross-references and links between contents. Thus, in addition to redundancy, a variety of searching pathways can be discovered. These pathways usually are unexpected and thus surprising for the user of the card index. In this respect, secondary memory enjoys not only a physical but also a cognitive?expansion. Memory functions in the manner of a historical machine in the cybernetic sense of the term. Every reaction of the machine is contingent on the past. Consequently, the relationship between input and output is not invariant but variable;14 it depends not only on the fact that the user will discover new file cards when searching through the filing cabinet but also (and even more) on the fact that the structure itself may enlarge each time the user triggers the combinatory structure of his private archive and the number of possible combinations increases. Therefore, the card index is not simply a data storage system but a genuine structural device that (in addition to data) stores neighbourhood relationships, i.e., associative cross-references and links between data.15 This scenario has a twofold advantage. First, memorability undergoes exponential growth. Users can retrieve the same data through different exploratory routes, whereas the same data can become an opportunity to address a network of meaning associations that trigger latent data. The wider the network, the larger the retrievability.16 Second, the machine becomes unpredictable, and interaction with it has an information value for users. In any access, the users enter a maze that has changed in the interim. Thus, a search is no longer a means of discovering what exists (inventio) but is instead a genuine exploration (invention). The function of memory is not to produce the correct answer for any question as if it simply stored a list of records. Instead, the machine is built such that it changes after interaction with the user and stores the outcome of this interaction in the changed structure. The machine is user-adaptive and exploits this ability to enhance its inner structure.17 After each changing interaction, the machine reacts (even to the same promptings)?

through the changed structure. Since the machine learns from the users, the users can learn from the machine. Their interaction is a true communication process and involves two black boxes, both of which behave as Alter Egos with respect to one another.18

To initiate this singular learning system, users must trigger the web of references that constitutes the inner structure of secondary memory. The filing cabinet reacts to itself and – through such self-reaction – also to the user. This process occurs first by means of self-resistance. The issue is not simply to search the filing cabinet for what is there but also for what is not there. Meaning associations are never created at will, and neighbourhood relationships among data follow a logic although they remain contingent. Furthermore, the user must eliminate any possible inconsistency by carefully maintaining the catalogue. The same entry cannot be tagged with two different numbers, and two different entries cannot have the same number. Resistance also occurs when an entry comes to nothing or (even worse) when an entry is missing. In any case, resistance is always inside, not outside; resistance is in the system and not in the environment.

Self-resistance is combined with joining capacity. In this case, the question is how to reproduce variety through selection.19 Like every self-referential system, memory is not simply a store of data. Adding new elements exponentially increases the number of connections. Thus, a single element can produce an overload of connecting possibilities (which is the meaning of complexity). Selection results in a variety that requires additional selection. One filing slip is followed by another filing slip in the same entry, or a slip refers to other excerpts stored under different entries, which subsequently unfolds into another branching off of possible references. Once the network of references is triggered, the filing cabinet reproduces not only connection but also connectibility. In other words, in principle, every excerpt is possessed of the possibility of unlimited meaning associations. Thus, the joining capacity is not simply repetition of the same.

Florilegia contained large collections of quotations organized into commonplaces. However, such anthologies were not conceived of as a means of searching for novelties. By contrast, while users are coping with secondary?memory, the system itself reproduces elements through systemic elements and reproduces reproducibility. In other words, memory functions as an autopoietic system.20 Reproduction is self-reproduction because the combinatory performances that the card index can produce when its structure of references is triggered enjoy autonomy.21 The outcome is surprising because the structure is recursive. Every new entry and every new link affects the horizon of variety, i.e., the background against which data become information. The content of the filing cabinet may be the same, but its information value changes because – in addition to the variety that the user must address – the selectivity of the selection changes. How was such an improbable evolutionary advance possible?

6.2 Evolution as Deviation-Amplifying Process

The triggering point of every evolutionary advance is deviation. Without deviation, reality would always remain the same. However, deviation alone is insufficient. Systems can determine whether to accept or reject differences. Compared with rejection, acceptance is highly improbable. The theory of evolution aims at explaining how it is possible that the high improbability of a process getting started turns into the high probability of that process maintaining itself.22 From the perspective of social memory, deviation means forgetting learning. In short, evolution can be described as a process that fosters forgetting through communication memory becoming increasingly autonomous from the memory of individual consciousnesses.23 The social theory hypothesis posits that during early modernity the printing industry provided this autonomy the impetus it required to take off – an outcome that writing alone did not achieve because it was subordinate to orality in knowledge management. In the literature on the art of note-taking that spread in Europe between the?

seventeenth and the first half of the eighteenth centuries, there is substantial and compelling evidence of this structural change. There is a shared awareness that to be deviant is for many reasons much more suitable than to abide by the rules.

The first and perhaps most striking shift in thinking was the strong opposition exhibited by scholars to the use of the imagination to remember. Against the background of the art of recollection and its success until early modernity, such an opposition appears nearly irreverent. In the mid-sixteenth century, scholars realized that using the imagination for knowledge management was childish and even ridiculous. According to Drexel, to stock the storehouse of memory with a copious volume of images for later retrieval not only required great effort but also wasted time because images escape from this storehouse as prisoners escape from a jail without guards.24 Students who wished to entrust their recollections to reliable keepers – Drexel recommended – should make annotations (“notas & excerpta”).

In fact, nearly a century earlier, Agrippa von Nettesheim had written a savage criticism of the art of remembrance as taught by Cicero and Quintilian. Artificial memory burdens “natural memory with infinite images of things and words”. Thus, it “methodically drives those who are not satisfied with natural limits mad”.25 Compared with Agrippa’s criticism, the attitude of later scholars remained the same. According to Alexandre Fichet, trusting psychic memory was like trying to drink from a running river using a sieve or entrusting one’s treasure (literally one’s storehouse) to a thief.26

The opposition to this system of mental knowledge retrieval peaked at the end of the seventeenth century. The striking fact is that this opposition was always combined with a comparison with the clear advantages that the art of note-taking offered scholars and that were only then becoming apparent. For instance, in discussing artificial memory, Lodovico Antonio Muratori argued that – compared with the art of note-taking (trascegliere e notare) – the technique employed for recollecting had little use and should be abandoned. According to Muratori, the absurdity of this technique was that it doubled references. In addition to the memory required to store images in a fixed structure of places, another memory was required to recollect all meaning associations (applicazioni) that coupled memorable things with images,27 thus leading to an endless regression.

However, memory is always contingent upon a doubling of references, and the way in which this doubling is put into practice reveals the type of social memory that scholars are dealing with. Doubling is required because if it were possible to retain the presence of everything that must be remembered (as autumn by Aristotle), nothing would be forgotten. Therefore, there would be no need to create a system for recollection. Instead, by doubling references, it is possible to combine self- and hetero-references, and as with every technique, scholars must train this combination of references to be able to achieve a good performance. Thus, Muratori’s objection is somewhat na?ve. But it becomes reasonable against the background of an option that previously was unavailable. This option consisted of entrusting recollections to paper – again, a deviation from the transmitted rhetorical culture.

The Renaissance revival of the commonplace-book was based on the firm belief that paper was a more reliable mnemonic device for knowledge retrieval. According to Drexel – a learned man with a deeply ingrained typographic mentality –, there was a circular relationship between paper and memory. Because knowledge was increasingly retrieved from books, it was to books that scholars should entrust all the knowledge they did not want to forget.28

The novelty of this opinion can be grasped against the background of Plato’s opposition to the use of writing to manage knowledge. Plato was genuinely frightened by the prospect of what we calmly regard now as an evolutionary advance. In particular, Plato was afraid that those scholars who had acquainted?

themselves with the use of writing to recollect scholarly knowledge were placing increasingly more trust in an external support. On the one hand, this reliance on writing implied a loss of control over the circulation of texts because, once it is published, a text becomes independent of its author. On the other hand, learned men had become dependent on the texts to which they had to resort to remember what they had neglected to fix in their minds.29 The decay of society that concerned Plato implied not only a reflection on the ambivalent relationship between writing and forgetting but also the feeling that knowledge could be autonomously managed through the recursiveness of texts rather than through oral communication – an outcome of the increasing differentiation of interaction and society that only the advent of the printing press would definitively effect.

Renaissance scholars who encouraged the recourse to paper (obviously also in a metonymic sense) were aware that Plato’s objections might be exploited by opponents to challenge the usefulness of the art of note-taking because paper as a hypomnematic device was a good alibi to neglect learning.30 Nonetheless, they chose paper. For instance, Muratori argued that paper is a stable external memory which reason has to consult in order to retrieve with more certainty what gifted men find inside themselves – although less quickly.31 And nearly twenty-five years earlier, returning to Drexel, Andreas M. Stübel distinguished between psychic memory, which is a type of primary paper, and the commonplace-book, which is, in contrast, a secondary and subsidiary memory (memoria secundaria & subsidiaria) to which scholars should entrust their knowledge.32 Evidently, Muratori and Stübel distrusted personal memory – it is unstable compared with books – and re-assessed note-taking as an external memory aid. They shared the opinion that the student who makes excerpts does not repudiate memory’s usefulness but simply considers excerpts to be a more effective device (adminiculum) against forgetting.

The evolutionary advance became clear when deviation was widely accepted and consequently made subject of publications. This transition occurred between the first and second halves of the seventeenth century. Perhaps no one expressed the preference for deviation so convincingly as Drexel. According to the German Jesuit, teachers of rhetoric were foolish, and the use of a virtual space to fill up with images that act as mnemonic hooks was a waste of time. In addition, every technique whose purpose was to train psychic memory was self-defeating. Drexel argued that his Goldmine aimed to teach how to make excerpts, not how to remember, and he implicitly meant that, while reading his book, the reader could learn how to forget.33

However, such evolutionary advance did not occur instantaneously. Even the most fervent supporters of commonplace-books had hesitations and second thoughts. In a sense, the selection of deviation implied a performative paradox. In other words, long after printing had become an industry, contemporaries still had only an outdated language to describe a new habit. Weighing pros and cons, they oscillated between excitement regarding the cognitive advantages offered by the evolutionary advance and scruples prompted by the cognitive habits transmitted by a still authoritative tradition. In fact, the book market enabled one to develop a good command of classical and medieval culture at little expense, but it also tempted learned men with a plethora of discoveries and new disciplines. Thus, considering the effects of his teaching, Francesco Sacchini sensed a danger in the use of secondary memories: the atrophy of personal memory. Therefore, he advised students not only to make excerpts but also to learn annotations by rote to prevent memory from becoming lazy. That he was teaching to file, he added, did not mean that he considered mnemonic training to be negligible.34

When Drexel, who availed himself of this reasoning, similarly stated that it does not suffice to make excerpts but that students must also repeatedly read the excerpts and fix them in their minds,35 he did not realize his advice to be a contradiction. If the duty of teachers was not to teach how to recollect?

(meminisse), they should not advise students to learn by rote (ediscere) what they annotated. This paradox emerges even more clearly when Drexel observes that it does not suffice to write down annotations but that students must also remember whether and what they annotated.36 They were in fact compelled to recall what they tried to forget while annotating.

This genuine puzzle derives from the fact that an individual who trains a secondary memory abandons training his personal memory. Sacchini himself was well aware that the learned man worthy of this epithet should not stock his library with books but his mind with cognitions, and he should not stuff his commonplace-books with excerpts but note what is memorable in his soul. Nevertheless, if he taught to take notes, the reason was that memory far from being neglected – was substantially more effective (felicius), as Drexel also noted.37 First, because note-taking prompts the reader’s attention. Consequently, the reader reflects longer on what he is reading, and the matter becomes more clearly understood. Second, what is read can be better fixed in the mind because note-taking compels the reader to pause over the text, to re-read it, and to engage with it more thoroughly than he otherwise might have. Third, for all these reasons, excerpts and annotations represent a highly effective remedy for oblivion.38

The question could be raised as to who was right? Those who believed that the habit of note-taking makes memory lazy – in this respect, Francis Bacon stated that knowledge can be stored either in writing or in memory and that although a good digest of commonplaces is highly useful, one must also remember that commonplace-books cause “a retardation of reading, and some sloth or relaxation of memory”39 – or those who stated that the same habit strengthens memory, instead? The answer is both. Indeed, when one becomes accustomed to using a filing cabinet, what is atrophied is simply the psychic memory, not the social memory. In fact, the latter is strengthened. Like any secondary memory, the filing cabinet enables society to remember substantially more than was previously possible because it enables users to forget substantially more.40 It is somewhat reasonable that this oddity was not?immediately understood by contemporaries. For some time, the situation remained, as élisabeth Décultot aptly stated, “incontestablement équivoque”.41 Thus, although scholars believed that they were entrusting knowledge to the filing cabinet to remember it better, they gradually and irreversibly became accustomed to forgetting.

6.3 Storing Expansions

Selecting deviation is not enough. An evolutionary advance occurs when it becomes clear that deviation presents so many advantages that only deviant individuals would abide by the old rules. Gradually, deviation becomes so normal that individuals simply disregard what was previously novel. Society adapts to the cognitive opportunities that it itself makes available, which represents a type of systemic self-adaptation rather than an adaptation of the system to the environment. In this respect, evolutionary theory speaks of ‘re-stabilization’. Because the evolutionary advance that is the subject of this essay concerns knowledge management, it is reasonable to ask what re-stabilization consists of when scholars make use of secondary memories. However, investigation becomes more difficult because re-stabilization attracted less attention than deviation and was only occasionally addressed by contemporaries.

Social knowledge management always implies a special circularity that is contingent on available media. The evolution of such circularity hides in metaphors by means of which society represents to itself its own relationship with knowledge. Semantics furnishes plenty of empirical cases. For example, in early modern Europe the word arca suddenly ceased to refer to the rhetorical storehouse and became a synonym for secondary memory. Similarly, knowledge was unexpectedly conceived of as a systema, a term that previously (although incidentally) existed in the Greek philosophical language, disappeared during the Middle Ages, and suddenly re-emerged at the beginning of the seventeenth century as a keyword in the title of philosophical and theological handbooks.42?

This type of semantic shifting may be understood as compelling evidence that cognitive habits were changing, although no contemporary could yet explain how.

Novelties in the methodology of studies and scientific research compelled early modern scholars to re-assess the form and function of the book. An idea that began to strike scholars during the early modern period was that as books are required to feed a filing cabinet, so a filing cabinet is required to publish books. During the same period, the format of the book underwent a substantial transformation. On one side, commonplace-books were deemed a substitute for libraries.43 On the other side, books were increasingly edited to enable their usefulness as card indexes. They were equipped with large subject indices, and this feature soon became a marketing device (i.e., a form of advertising) behind which was hidden an awareness that an index fulfilled reader expectations. Scholars perceived the book less as a repository of memory and more as a type of bookkeeping system that might be consulted as the need occurred. As a card index, a book should be readable in a highly selective manner, such that each reader might obtain information per se. In short, the book was understood as an ouvrage de référence that was made not to be read from the first to the last page but to be consulted per intervalla (by skipping pages), as Conrad Gessner noted in the mid-sixteenth century.44

When scholars understood that they could entrust everything to their own filing cabinet, there was no reason not to read and select everything memorable they found in books. This experience intensified in a period when the typographic industry was whetting reader appetites that the industry itself aimed to satisfy with the continuous production of books. Thus, when Drexel boasted that students could read one hundred or six hundred authors in different languages in a day and in whatever sequence if they used his annotation?

system (enotandi methodus), his hyperbole clearly displays the feeling of freedom from the task of mnemonic reading that drove early modern scholars.45

The habit of excerpting inverted another essential rule. In the rhetorical culture, reading too many books was considered to be a vice. Until the end of the Middle Ages, the prevailing notion was that an individual who reads too many books cannot store them in his mind, as someone who eats too much ends up vomiting. Consequently, scholars were supposed to acquaint themselves with few authorities. The common view was that someone who reads all he can is akin to a vagrant who finds many hosts but no friends.46 Reading should not be agitated and disorganized. As Petrarca posited, scholars must stock their memories – and not their libraries – with books; thus, knowledge must be stored in the mind, not filed away on bookshelves. From the copiousness of books (copia), only aversion and laziness (inopia) result.47

Two centuries after the invention of the printing press, in a culture that had definitively absorbed the typographic mentality, the habits of scholars radically changed. In the mid-seventeenth century, John Locke stated that a learned man is a ‘a(chǎn) bookish one’. A century later, Johann Andreas Fabricius reinforced the notion that it was impossible to become a learned man without reading a large volume of books (“Man mu? nicht ohne Bücher wollen gelehrt werden”).48 And since there was no limit to the production of new books, a type of memory was required that could provide unlimited storage.

In this respect, the evolutionary advantage of the filing cabinet becomes clear. What actually changes during the first two centuries after the invention of printing technology is the function of commonplaces. In the rhetorical culture, according to Quintilian’s standard definition, commonplaces were ‘seats?

of arguments’, and topics were stored there in preparation for later retrieval.49 This notion of places was practically managed through multi-chambered virtual constructions which could be supplied with images of memorable subjects. One of the essential rules of the construction of these buildings was that they should be neither too large nor too small. If they were too large, the orator might lose his bearings (i.e., he would forget). Should they be too small, there would not be sufficient space to store everything.50

Between the sixteenth and the seventeenth centuries, this idea of artificial memory underwent a radical change. Instead of training in walking through the storehouse, learned men compiled commonplace-books to be used – as previously discussed – no longer as mnemonic aid but as a secondary memory, a ‘stable external memory’, as Muratori called it, which every user could browse through to obtain information. One of the most striking effect of this transformation was that memory no longer had limits in a physical and in a structural sense, as well. Topics were no longer understood as a place for storing latency but as “a place for storing dilations and expansions of a theme”.51 As a consequence, the primary concern was no longer with training individual memory but instead with institutionalizing the advancement of learning.

To achieve this result, scholars had to abandon the certainty of permanent knowledge to which additional contributions could be made without changing the geographical order of topics, and to accept the odd, initially striking idea that knowledge is transitory and that it ages. The regular practice of annotation supported this transition by triggering a causal loop of positive feedback. In the era of the printing press, reading does not entail gathering all the knowledge that is worth being remembered but searching for interesting facts, hunting for information. Readers no longer aim to remember but to expand their filing cabinets by adding new file cards and entries that they can freely recombine. Thus, Renaissance education based on the compilation of commonplace-books changed intellectual priorities and emphasized the dismemberment of knowledge into loose entries. If students had to select commonplaces with which to compile their copybooks while reading, they also approached knowledge with dismemberment purposes, so to speak, by looking at previously compiled commonplaces. The unanticipated consequence of this approach was a type of knowledge self-reaction and the arising of combinatory habits that led to unexpected cognitive opportunities.

‘Storing expansions’ may be understood with double meaning. On one side, the goal is to store the continual knowledge expansion enabled by the typographic industry. On the other side, the goal is to expand the storage potential of secondary memories. The card index fulfils both duties, and it is ‘futurecentric’ in this sense.52 Like archives and libraries, the card index is oriented towards an open future. Each new entry (or file card) provides memory with unexplored relationships and meaning associations, and every expansion reproduces additional expansibility, as a consequence.

This outcome is precisely what the rhetorical storehouse could not afford. As large as it was, it remained a closed space. Only in this way could an orator orient himself, that is remember. If he wished to add new recollections or to replace previous recollections, he had to destroy many images stored in the virtual buildings of his artificial memory. This activity was exhausting because what the orator had fixed in his mind could only be erased with difficulty. By contrast, the loosening of knowledge into elementary units and the careful maintenance of the filing system offered an unusual freedom (compared with the past). In principle, everything could be placed into the card index without regard for the consistency of the content. The primary condition was that every entry was linked to the network of references of which the memory structure consisted. The check of consistency was no longer contingent on space (the intellectual activity of medieval scholars was in fact highly similar to a local movement) but on catalogue. For the same reason, the card index was “supremely tolerant of cognitive dissonance”.53

The function of commonplace-books also changed. Compared with topics, whose function was to store redundancy and keep it handy, early modern commonplace-books were used to reproduce variety. In the former case, the orator consulted the book to retrieve a known matter that he had forgotten over time. In the latter case, the user exploited the combinatory craft of the filing cabinet to determine what was stored and then to search for novelties. In other words, early modern commonplace-books functioned as engines of variety, not as engines of copy.54

The counter-intuitive outcome is that selections made through annotations were not only a means of avoiding too much variety (i.e., too much to know) but were also a requisite for reproducing substantially more variety than before. This double performance explains Drexel’s paradoxical statement in which he emphasized that Justus Lipsius had been able to develop his large and distinguished erudition (“tam copiosa(m) et illustr(em) eruditio(nem)”) and to edit so many books (“tot librorum fecundita(tem)”) because he had not only read a great deal but also employed a filing system. Thus, the copiousness of his scholarly production – a common feature of learned men who used card indexing systems – resulted because he learnt by selecting and making excerpts (“seligendo et excerpendo”).55

From an evolutionary perspective, this situation could be re-described by stating that in early modern Europe, social knowledge management was restabilized by selecting a continual (one could also say methodical) reproduction of variety. Perhaps the most striking aspect of the evolutionary process that resulted in the construction of secondary memories is this form of dynamic stability.56 However, the advent of typographic technology alone is insufficient to explain this evolutionary advance. The dynamic side of stability depends on the reproduction of new knowledge. In turn, this reproduction is the function of a system, that of science, which did not develop autonomy until early modernity. This change of social structures is the topic of the next chapter.