Second edition
Abstract: This essay explores the evolving role of historians, philosophers, and social scientists in the context of the network age, characterized by rapid technological advancements and societal transformations. It analyzes the impact of AI, algorithms, and digitization on historical research methodologies, societal narratives, and ethical considerations. The essay highlights the need for interdisciplinary engagement and critical reflection to address challenges such as historical misinformation, societal polarization, and the manipulation of information networks. It calls for historians to actively participate in shaping public discourse, combatting misinformation, and preserving the integrity of historical narratives in the digital era. Through a comprehensive examination of contemporary issues, the essay underscores the relevance of historical analysis in understanding and navigating the complexities of the network age.
Recently, the network era can be considered a new paradigm or at least a new era. The breathtaking numbers of social and professional networks, the speed at which they generate, consume and disseminate information, with the new technological developments of algorithms and AI of the last decades, represent a significant change for the practice of history. These deserve more attention.
In the late twentieth century, in order to make the most of the Internet’s immense potential, industrialists, scientists and technocrats devised ever more complex new ways to amplify the power of computers. During that quest, the algorithms used to achieve that new goal themselves became increasingly complex. During that quest, the algorithms used to achieve that new goal were themselves becoming increasingly complex.
Some historians in the early twenty-first century were also tempted to exploit the richness of those algorithms. Data mining was applied to analyze a large amount of texts with innovative results. Those algorithms could be the answer to the revolutionary changes and challenges posed by societal digitization. While yesterday historians studied the few relics left to us by time, tomorrow they will be overwhelmed by the plethora of sources provided by digitization.
This opinion piece seeks to put that disruptive change with its challenges on the map. It urges humanities to recognize, understand and meet those new challenges.
Challenges such as the abundance of sources, or a methodological procedure for online primary sources that must be processed according to a yet to be determined criteria to become usable material for research, can be divided into two larger categories. The first is ontological, the second is societal. First, the ontological issue of rapidly changing AI and algorithms. The methodology, la raison d’être (the reason for existence) and ethics. The second category is dealt with briefly here and will receive more attention later, in another paper.
Raison d’être?
Starting with the raison d’être. Below is a brief explanation of the raison d’être of the practice of history. Here it is necessary to keep two terms well apart, historical awareness and social memory.
Historical awareness is one of the fundamental qualifications of the historian, and of human beings in general. It is a measurable psychological and social variable, since not only people but societies, institutions and organizations of all kinds possess some degree of it. It is not based on a fixed, objective memory. As historian John Toch emphasizes, societies and people forget parts of history, cover them with other memories or emphasize other parts of history.[1] With the forgetting and changing of this history, the need for a search for objectivity of the past, a goal of returning the past to its truest form, has always been the historian’s mission.
Social memory and historical awareness share characteristics and can be confused with each other while there is indeed a difference between the two. When one speaks of historical awareness, one speaks of a rigorous interpretation of History. Social memory, on the other hand, is based on a less rigorous historical basis. [2]
Social memory, as the name implies, is the memory of a group, society or population on a highly selective historical basis, due to the normal human instinct for social gathering and a sense of belonging based on cultural, social and historical aspects. Social memory is the creation and basis of self-identity and feeds off the historian’s historical findings.
Historical awareness and social memory are social and psychological variables that attempt to answer certain ontological questions of existence: where do we come from and who are we? One of the historian’s tasks is to interpret the sources, relics of the past that have been handed down to us, in an attempt to answer these questions as close to historical truth as possible. The historian does this by interpreting the sources and artifacts, by giving a new reading of the sources, by putting forward a new theory or revising an old theory; and through academic debate, historical science arrives at a historical consensus: a truth that comes closest to the unattainable historical truth.
Do new technological disruptions change this goal? Is this reasoning still relevant, given that an artificial intelligence and an algorithm have infinite memory and millions, if not billions, of sources at their disposal, with increasing interpretive power over time? Notwithstanding the fact that in the near future historical analysis of society will not be possible without the help of computers because of the vast amount of resources, or data, created by those same artificial intelligences and algorithms and their users.
One aspect of the answer to the question of whether technological change nullifies the historian’s raison d’être is twofold. On the one hand, a computer is not omniscient: it is always possible for data to be misinterpreted due to a bug in the system or a lack of calibration or knowledge. Moreover, algorithms for cross-referencing between sources have so far been created mainly for computer-generated sources, which is technologically much less complex than for handwritten historical sources. Thus, the historian is still needed, at least to control the historical data output and interpretations of these programs.
Let us delve into history in search of support for the claim that the historian is still needed. In his never-completed theory of linguistic structuralism, the nineteenth-century linguist Ferdinand de Saussure distinguished the binary characteristic of the word or sign. The signifier is the concretely realized sign, the external form. The signifier is the mental (human) concept to which the signifier refers. Thus, if we take Saussure’s thought into account, it will be noted that a computer is not yet able to understand a word as a human does. The signifier (signifiant) and significate (signifié) of words differ according to culture, norms, the author’s command of language and social conventions.[3]
Taking these aspects of analysis into account, this is still beyond the capacity of AI. The resulting machine interpretation is likely to be less accurate and less close to historical truth than a historian’s analysis in this sense. Of course, it is also worth noting that man himself is time-and-place specific, with his own cultural, social and intellectual background, so he himself cannot interpret everything or do so incorrectly either. The difference is that the historian is, or should be, aware of this, and thus can get closer to historical truth than a machine. The reason why a historian can if he is historically aware can be found in the philosophy of Gadamer, German philosopher of the twentieth century.
Since that Understanding involves reaching agreement through which otherwise incomprehension imposes itself, Gadamer argues that Understanding involves something like a common language, albeit a common language that is itself formed in the process of understanding itself.[4] The common language is constantly formed and reformed in that same process.[5] In this sense, according to Gadamer, every understanding is interpretive, and insofar as every interpretation involves the exchange between the familiar and the foreign, every interpretation is also translational. [6]
Applied to history, this could mean that even though the historian is place- and time-bound and culturally bound, through a hermeneutic conversation with (sources from) the past, he can get interpretatively close to historical truth because he then has a (better) command of the common language.
However, that the resulting machine interpretation is likely to be less accurate and less close to historical truth than a historian’s analysis is not a given in the future. Nothing can assure us that this will not be within the reach of AI and algorithms within 20 years. When historians today are already preaching the utility of algorithms and programs that enable data mining, text analysis and text mining, and showing the innovative insights that such analyses can provide historians, two questions arise: will the historian’s purpose change? And second, what would be the implications of AI-constructed historical information as the basis for social memory and historical consciousness?
Raison d’être, social memory and society
For yes, this is another aspect of the answer to the question of whether technological change changes the historian’s raison d’être. If ChatGPT is powered by other algorithms that transcribe historical sources as pre-work, ChatGPT could generate innovative historical essays or perhaps (let’s be imaginative) new historical information and theories. What does this mean for social memory?
A misinterpretation, a bug, or wrong information may still emerge with the algorithm. This will impact to social memory, and thus society in the future.
In the future, will the historian be a checker of the various interpretations created by the AI through its sources-crossing algorithms? It is possible that he or she will investigate the AI’s interpretations the same way a detective does: closing improbable paths and doors, with insufficiently proven arguments, with false or misinterpreted facts. This is certainly a possibility when it comes to the problem of how many sources historians will have to deal with in the future. In that case, there will be a shift in historiography: from “maker” or “discoverer” of the past to controller and thus a shift in emphasis in the historian’s raison d’être.
Technological change also means that historiographic methodology changes. The evolution or change of methods within historiography is not new, since historians always keep (and have kept) a critical eye on how the unattainable historical truth might be discovered, or at least how it might be approached as closely as possible. As in the natural sciences, Popper’s theory of falsification can provide the basis for historical research, perhaps in a different form. While historians, as highlighted earlier in this text, interpret sources and, by giving a new reading of the sources, put forward a new theory or revise an old theory; in short, put forward a possible falsification of a theory; historical science thus arrives at a historical consensus through academic debate.
AI and algorithms will not change the basic scientific theoretical methodology of historical research, but they will in practice. Indeed, a significant shift is taking place within historiography. While historians are used to working with few sources (the scarcity of sources increases as one goes further back in time), a radical change in quantity is taking place with the development of digitization. The digitization of government, business and organizational data is growing, increasing the available resources exponentially. AI and algorithms are pushing this phenomenon into another dimension, as it is now possible to collect data and personal resources about each individual; getting to know them personally on a never-before-seen personal level, with the help of programs and using AI or algorithms, creating an even more significant amount of data. A substantial amount of data that grows at frightening speed as AI and algorithms learn more and more about humanity. Should the historian also become a computer scientist?
The question can be asked (again), because it seems that historiography has not yet asked it seriously; courses at universities hardly address the subject. A reflection within historiography on research methodology and the use of digital tools should be high on the agenda.
Ethical Questions of History
Finally, with technological change come ethical questions. There is no denying that historical processes accelerate exponentially over the centuries. After the atomic age, the information age, it can be said that we are now in the network age, and that in little more than half a decade. Social and technological changes are the causes of this historical change. For example, historians and journalists are already trying to analyze the capitol storming in Washington in 2021. This may pose an ethical problem, however, since historians are still debating their role in relation to the present. Yet the question is not new. Technological change is intrinsically part of history and time. There is no denying that history was made at the time about technological changes and their impact on society. But was it reported by journalists or historians? As technological, scientific and social change accelerates, how do you really make a difference? As the past comes closer to the present, what should the historian do?
Another ethical question, this time more practical. When studying the Jan. 6 uprising on Capitol Hill, for example, the actors who participated in that historic event are still alive. That means they have a right to privacy, even if the data generated by AI and algorithms are technically capable of telling with some degree of reliability how and why that event took place. The (historical?) data is already there. How will the historian position himself toward questions of privacy when society undergoes events that form identities and create social memory in a very short period of time in an accelerated historical process?
Social memory and implications of algorithms and AI
Finally, I want to talk about social memory and the impact of algorithms and AI on it; and the role of historians vis-à-vis this development. Algorithms and AI have brought with them a new era. An era of networks. The network era is characterized by the ease with which networks can be formed; they are formed by social, ethnic, cultural and intellectual groups/connections with (trans)national dimensions. It is easy to find a like-minded person or someone with similar tastes or interests on the other side of the world and thus become part of a network or start one. Forums, think tanks or cultural and social reunions are as numerous as they are easy to set up thanks to the Internet and provide a broad platform for expression. These networks, with primary cultural, social, economic or political motives, transmit information at lightning speed in anarchic and synergistic ways.
This way of sharing information also applies to information created, published and shared around history. Fake news about history is now commonplace, and historical conspiracy theories are making a comeback, not to mention growing historical revisionism. This new development in the network age is crucial for the humanities and history in particular. As we saw at the beginning of this essay, historical information is the foundation upon which social identity is built. History can be seen, felt and experienced in any society and any (inter)national social encounter. When misinformation about history is written, published and shared, it undermines this identity. Revisionism, historical disinformation and historical conspiracies existed in the past but these new forms of revisionism differ for two main reasons. The first is the speed and scale of the information sharing process; the second is the impact it can have on individuals and society in a short period of time. It leads to an unprecedented polarization of society, of identity. People who are in their own information network almost all the time see their knowledge and beliefs change. In the case of history, this can be dramatic because of its intimate connection to identity.
Here we return to the previously mentioned comment about ChatGDP. If an AI such as ChatGDP, which extracts subjective information from the Internet, writes history, shaping the identity and reality of the past based on a one-sided selection of direct and indirect sources (including fake news, manipulated history/information) and shared in dizzying ways, across thousands of information networks, how should historians – the scholars of the past – proceed? And should historians actively fight against this new form of misinformation?
Conclusion
For the historian, it is important that historical knowledge remain as close to historical truth as possible. We must therefore reflect on the just-named new challenge as it affects society in unprecedented ways. A written historical narrative can have great consequences for the future, as nineteenth-century nationalist historians can attest. It is therefore important for historians to recognize the problem of the democratization of historical knowledge and revisionism resulting from the changes brought about by the network age.
This text has asked many questions and given few solutions or answers. This is for the simple reason that those answers have not yet been discovered, at least by me. Historians are still relevant, and in my opinion, indispensable to society. One piste for an answer to the questions formulated is that the historical must take his social role to heart. This involves the historian’s participation in the debate. His or her expertise in text and source analysis and historical methodology is, in the network age, of the utmost importance. This involves explaining history, but also and above all bringing history and the general public together to prevent or at least combat historical disinformation or the misuse of historical information.
Given the intimate connection to identity and the historian’s at least partial involvement in the creation of identity, as well as recent social developments, it is crucial to think about the questions posed by algorithms and AI.
[1] John Tosh, The Pursuit of History: Aims, Methods and New Directions in the Study of History, Sixth edition (London: Routledge/Taylor & Francis Group, 2015), 2-35.
[2] John Tosh, The Pursuit of History: Aims, Methods and New Directions in the Study of History, Sixth edition (London: Routledge/Taylor & Francis Group, 2015), 2-35.
[3] Ferdinand de Saussure et al, Cours de linguistique générale, Grande bibliothèque Payot (Paris: Payot, 1995).
[4] Jeff Malpas, The {Stanford} Encyclopedia of Philosophy, ed. Edward Zalta and Uri Nodelman, Winter 2022 (Metaphysics Research Lab, Stanford University, 2022), https://plato.stanford.edu/archives/win2022/entries/gadamer/.
[5] Jeff Malpas, The {Stanford} Encyclopedia of Philosophy, ed.
[6] Ibidem.
Leave a Reply