In the last weeks before the American presidential election, like most of the public I suppose, I believed the polls and statistical models that predicted Hillary Clinton would win. Belief was the only thing left to me, for one simple reason: I am a mathematical and statistical ignoramus, even though I am about to finish a master’s course at the oldest Czech university, and so one might expect that, as a member of the academic community and of the notional intellectual class, I would be properly equipped for my future life with this knowledge too. Unfortunately, that is not the case. And I blame no one but myself. I have thought about my inadequate mathematical education several times before, and much earlier, but it is precisely the US presidential election, or rather its result, that makes me reflect on what it means today to be an intellectual and a pseudo-intellectual, an educated and an uneducated person.
The field I am currently studying has, among other things, shown me relevant literature from which it is evident, under the veil of big words and long sentences, how technology and technical thinking dominate today’s world. Although we hardly notice it in our everyday activities, most of the work, development and research that has the greatest practical impact on society is at its core the result of technical thinking, or more specifically a result based on a theory whose most precise expression takes a mathematical form.
And this applies to today’s humanities as well, which have for some time now been reluctantly preparing for the fact that verbal abilities will not be enough for future academics: bad research in the humanities is being replaced by good statistical analysis, whereas good research in the humanities can hardly do without algorithmic analysis. It may seem that, for example, studying literature does not require a technical and mathematical education; quite the opposite, people often study literature precisely because they do not have such an education and do not want it. That assumption holds. Getting a degree in this field does not require the student to be equipped with programming skills, but if a person wants to embark on a professional career as a literary scholar, the situation is not so obvious. On the contrary, in recent years it has turned out that the use of sophisticated data analysis brings genuinely new scientific findings. For example, we could recently read that a computer analysis of the style of Shakespeare’s plays attributed co-authorship of the play Henry VI to the Bard of Stratford’s eternal rival, Christopher Marlowe, who had long been considered one of the candidates in alternative theories of the authorship of Shakespeare’s work. The computer analysis did refute the idea that Marlowe was Shakespeare himself (or the other way round, if you like), but a recognised co-authorship of one of the most essential works of the traditional Western canon is not insignificant. Both for Marlowe and for the future relationship between the humanities and computational methods, which in academic circles is called, somewhat vaguely, Digital Humanities. This relationship, however, confirms what has already been said: computing power and clever software are changing and supplementing established methods in the humanities (and not only there), while it still holds that the new algorithmic methods succeed when they are accompanied by high-quality, as objective and precise as possible, classical research. One does not work without the other. At least not until artificial intelligence produces creative results comparable to human ones. Whether that will ever happen is uncertain. For now, computers serve scientists only as a cognitive prosthesis; the human element is still indispensable in many cases.
Because we look at the world more (in the social sciences and the humanities) or less (in the hard sciences) from the position of a human observer, we must reckon with the fact that the social world will always appear to us through our human errors, cognitive biases, ideologies, prejudices and lack of knowledge. We often say that an uneducated person is more susceptible to opinions based on emotions, and to a tendency towards irrational reasoning in general. But there is a problem on the other side too. The more educated a person is, the more his life and career are invested in a particular, and it must be said narrow, part of the views of the world. Where the “uneducated” man merely speculates with his mates over an evening beer and in the morning goes to a job that has nothing to do with politics, the “educated” man defends values to which he devotes his whole professional and often private life too, because he has to. How could he undermine the values that secure him an income and social prestige, not to mention that most of his social contacts consist of people like himself? From this it follows, for me, that the educated man may paradoxically be much more prone to hasty and emotional reasoning, because unlike the uneducated man, who at most embarrasses himself among friends over a beer, the educated man’s life collapses if it turns out that he has been wrong all along. Being wrong means an existential threat to him. That is why the advocates of Stalinism, Maoism, behaviourism and other thoroughly unscientific theories defended these ideas to the very last minute, even after it was clear what atrocities they had committed on people and minds.
So there are the intellectuals, from whom dogmatic clinging to their opinions can be especially expected. Unfortunately for us, students official and eternal, it is precisely they whom we look up to with all too much trust as totems of true and undistorted knowledge. But it need not be so, if we are able to judge for ourselves whether their ideas are worth our time. For that, however, a student needs some already existing knowledge on the basis of which to evaluate new knowledge. We generally acquire knowledge through our own experiences, intuition, spiritual revelation and, last but not least, yes, you guessed it, again from the intellectuals already mentioned. And although there are those who can, in a dark corner of their room, without a supply of other people’s ideas, derive mathematical equations, understand Marxism, write Shakespeare’s sonnets or build a small nuclear reactor, most of us will pick up a book and absorb, line by line, someone else’s pre-chewed ideas, only to be able to reproduce them as our own, with or without a citation in a footnote. It seems that if we want to avoid uncritically accepting the ideas of one intellectual, we must accept the ideas of another! It is a vicious circle that we will not easily escape.
What we can do is come to terms with the fact that most of the knowledge we know is simply wrong or, at least, as my computer science teacher used to say, untrue. New truth is born at the margins of the mainstream of current knowledge, never at its centre, which mostly serves as the custodian of the status quo, that is, of the truths generally accepted in society, to which intellectuals turn as to the “consensus”, and people who have not the faintest idea what these temporary truths actually mean (i.e. journalists, writers, political commentators = verbal people without a mathematical and statistical education) as to “what the experts say”. But let us not despair: many people before us went through the same process of groping for the truth, and yet they reached some new knowledge.
We will not attain truth and objectivity any time soon — our brains were certainly not made for it, and it is something of an accident and a miracle that the gelatinous mass that once served for hunting mammoths and gathering berries can now calculate when the universe began — but we can approach them through the slow and careful work of accumulating information, sorting and comparing facts, to which we gradually add our own thoughts. Most of the time we will move in shadows, but now and then, at the periphery of our thinking, we will catch the light of information that bears some relation to the truth. As in the humanities, this activity was until recently a relatively successful procedure for what we generally call education. As in the humanities, much useful information could only be absorbed by verbal means. And for the same reason as in the humanities, this path is insufficient today.
After the election results were announced and the winner declared, I felt cheated. Not by the winning or the losing candidate, but by myself. Under the pressure of expert opinions and interpretations of research, I let myself be persuaded that the presidential candidate, Hillary Clinton, had a more than 80% chance of winning this election. In my defence, I was not the only one who succumbed to this “consensus”, and at the same time I was one of the many people who lacked the knowledge needed to at least verify the methods by which this consensus actually came about. I felt functionally illiterate.
This idea is not so far-fetched. Text, as a medium of thinking, serves more and more only for popularisation; the real core of knowledge is encoded in mathematical form. That, after all, is what Vilém Flusser wrote in his essay Alfanumerická společnost, in which he describes how text and traditional verbal literacy, once the preserve of the learned and the clergy, spread thanks to the democratisation of education; now, however, the alphanumeric society is governed by a new clergy that commands the language of mathematics (and, I dare add, the general language of algorithmic thinking and programming). This does not mean that the intellectuals of yesterday, such as the American journalist Malcolm Gladwell, will vanish at once. The function of popularisers of scientific knowledge will grow in importance for the lay public, because more and more of the public will (unless the current trend changes) be functionally or completely mathematically illiterate. For the new intellectuals, the guardians of knowledge, for mathematically minded people, Gladwell may be useless or even counterproductive, because he will be sowing plenty of untruths and lies in society.
Does this mean that everyone who wants to consider themselves an educated person today must necessarily study mathematics? It is quite obvious that in practice this need not and does not happen. But if the goal of an intellectual is truth (however naive this word sounds in a postmodern society), mathematical and statistical thinking is certainly a suitable tool for approaching the truth. This, after all, is understood even by the epicentre of all intellectualism, the universities, as they began to take mathematics into traditionally verbal disciplines. I have already mentioned Digital Humanities, the newest contribution of bringing mathematics into the humanities; similar revolutions could be observed in the mathematisation of economics, sociology and recently even history and theology, when the academic and historian Richard Carrier used Bayesian statistics to reassess the facts in studies of early Christianity and the existence of Jesus Christ.
The mathematisation of thinking, however, is not without its dangerous corners. Except perhaps for mathematics itself, where within some axiomatic system one can prove unambiguously that a given expression is true, or physics, where theoretical findings can be verified by experiment given the necessary technology, everywhere else one has to add some hypotheses of one’s own to the equations, and with them all that human baggage in the form of ideologies and cognitive biases. We thus arrive at the opposite extreme: mathematics “without lessons in sociology, history, cultural anthropology, ethnology or philosophy” can serve not only ideological propaganda but also destructive political and economic agendas. In the November issue of Literární noviny, Michal Rubáš gives in his column O ne-užitečnosti matematiky one of many examples: in the German economy it is calculated that “in the coming years it [will be] necessary to replace the loss of the dying-out domestic workforce with workers from abroad, and between the two variables they write ‘equals’. […] Had the managers of industry studied something other than economics or mathematics, they would have learned that even if many objects can be denoted by the same or similar signs, they do not thereby acquire the properties of the signs, nor do they begin to resemble one another more”. (Literární noviny, 11/2016, p. 2) Yes, even the queen of sciences, mathematics, can be grafted onto an ideology, especially when abstracting away all the real properties of the objects in the equations simplifies the calculations so splendidly, where a native and educated German can be exchanged for a person without knowledge of the local language, culture and society, a person without a background and without security in life.
What follows from all this? Verbal blather, without logical reasoning and without the necessary skills to obtain and machine-interpret Big Data; blather that in its argumentation targets the reader’s emotions, is just as toxic as the puritanical mathematical models of technical fachidiots. One has a diploma in rhetoric and English literature, the other in algebraic topology. These opposite poles met in the American presidential election, with the aim of unanimously installing in the presidential chair the first woman, a Democrat, the mother of all social warriors and the kindly grandmother of the big bad wolves of Wall Street.
Those functionally illiterate in statistics and mathematics could only look at a nice infographic that appears as seductive as a long essay in the New York Times full of quotations from literary works. Both outputs, which reduce complex data with many variables to a digestible output for the mathematically illiterate humanist majority, work as a meme that is (mis)used in society for argumentation, for intimidating political opponents and for disinformation. The mathematically illiterate humanist majority let itself be carried away by numbers that meant nothing to it. Journalists, political “pundits” and commentators spread and recycled this meme, which for them was a black box, epistemological fast food, because their education (and often their intelligence too) cannot digest anything numerical.
I feel that I was one of them. Not because I consciously spread their ideology, but because I could not come up with mathematical and statistical counterarguments that would express my disagreement with how the mainstream media interpret the (American) world. Without a deeper understanding of where the data come from and how they are handled, my role, for some the role of an intellectual, shrank to elegantly citing research that led me to a certain conviction. All wrong. I came a cropper like most people. When all the illusions that it is possible to outsource one’s decision-making to incompetent media icons have fallen, what next?
First of all, we need to keep our feet on the ground and say from the start that it is unrealistic for every true intellectual of the information age to download new data every day and analyse it overnight on their computer, only for everything to change the next day so that they could start all over again. More sophisticated algorithms, accessible to the public, can certainly simplify a good deal, but alongside such a solution my proposal is much more prosaic: the true intellectual of the information age should be able to take a raw collection of data and apply an analysis to it according to mathematical methods, but also according to hypotheses that he chooses himself, and he should be able himself to apply his hypotheses to a statistical model. Whether the resulting analysis has anything to do with reality will then depend on which hypotheses a person prefers according to his understanding of the world.
I therefore propose that everyone who considers themselves an intellectual, an analyst or another public figure speaking to the people should put their skin in the game: show us the data you start from; describe to us the methodology and hypotheses with which you interpret the data; finally, attach all this information to your article or publish it on your public profile, GitHub or the like. Argue (with) data. For the word is not an ideal tool for grasping dynamic complex systems. Enough of verbal rhetoric à la New York Times. Every university student must be able to download the data and the description of the methodology and run them through their own programs. If university is a preparation for life, numerical competence and dexterity with data are the Latin of our time; they are the lingua franca of the educated.
If someone longs for peace and happiness, let them have faith in others, even though they risk parroting pre-chewed food. Such a person, even a university graduate, will be cheap advertising and a useful idiot for all ideologues. But if they want to be a disciple of truth, let them search the universe of data.
(Send errors, typos and questions to hello(at)jakubferenc.cz)