Suicidal Empathy by Gad Saad

In a previous post I wrote about the book ‘The Parasitic Mind’ by Gad Saad and I mentioned that I saw my reading of that book as a preparation for reading his new book ‘Suicidal Empathy’. I have now finished reading the latter. I found it stylistically more pleasant than the previous book. Either his style has got milder or I have just got used to it. The most significant thing about the book is its contents. I think that it is a very important book and that it should be read by as many people as possible. One aspect of it is that the author presents many examples which are so extreme as to be breathtaking. Incredible extremes of wokeness. The main theme of the book is how Western society is being destroyed by excessive amounts of empathy applied to the wrong targets. The examples are backed up by sources where it is possible to learn details about the cases presented. I made use of this in a couple of cases. For instance I got more information about the case of the young German politician who initially lied to the police about the nationality of the men who had raped her so as to not risk arousing racist sentiments. Another important element of the book is its discussion of many themes from psychology, many of which were not familiar to me.

The first chapter is concerned with the notion of empathy in general and an initial presentation of its pathological variants. The second is concerned with things that people are not allowed to say in our society and even better (from the point of view of those promoting this ideology) not allowed to know. The central idea of third chapter is how people in our society make the fateful assumption that members of other cultures share certain of our basic principles. A key element of the fourth chapter is the way in which criminals are not held responsible for their acts, instead being portrayed as victims of circumstances. Subjects of the fifth chapter include COVID, climate change and trans activism. There I am less in agreement with the author than in other parts of the book, particularly concerning COVID. Chapter six treats topics related to DEI, or as Saad calls it, with deliberate misspelling, the DIE cult. It includes a quote which for me was one of the most shocking in the book, an oath taken by graduates of the University of Minnesota Medical School in 2026. This may be summarized by Saad’s title for the section: reject the hippocratic oath, choose the woke oath – be empathetic. Suddenly being well educated about medicine and doing everything to promote the health of the individual patient are qualities of a doctor which are of minor importance. In the seventh chapter there is a discussion of the way governments take the resources of those who are productive and distribute them to those who are not. The Canadian health service is taken as a central example.

When the central ills of our society have been presented the obvious question is what we can do to change this situation. The last chapter of the book makes some suggestions. When I ask myself what I can do I am disappointed by my lack of answers. I do not see this book as one which is to be read and then given a more or less honourable place on a shelf. I expect to be able to learn a lot more things by following the lead of the footnotes in this book. My recommendation is: buy this book, read it, recommend it whenever you get the chance. Its relevance is not confined to North America – it is relevant for all Western countries. As far as I am aware neither this book nor any other by Gad Saad has been translated into German. This is a pity. Presumably editors are afraid of touching an author with such a questionable reputation. By contrast, some of the books of Douglas Murray have been published in German. They appeared in the FinanzBuch Verlag and belong to a series run bei Tichys Einblick. Could that be a German home for Saad’s books?

Hartman’s theorem on conjugacy of saddles in two dimensions

In a previous post I discussed the topic of conjugacy of vector fields on , concentrating on the case where the vector field and the conjugation are smooth, i.e. . In what follows I will continue to consider smooth vector fields but allow the conjugations to have only finite differentiability. In the last post I showed that in the case all vector fields are smoothly conjugate near a non-degenerate zero. In particular in this case each vector field is conjugate to its linearization at a steady state. Consider now the case . The theorem of Hartman mentioned in the title of this post concerns a hyperbolic saddle in two dimensions. (The reference is ‘On local homeomorphisms of Euclidean space’, Bol. Soc. Mat. Mex. 5 (1960) 220–241.) In other words we consider the case that the linearization of a vector field at a steady state has one positive and one negative eigenvalue. The theorem says that the vector field is locally -conjugate to its linearization. This is a very special case as we can see by varying the assumptions. For instance, a general two-dimensional vector field is not -conjugate to its linearization. An example is the vector field which generates the following system of ODE: , . Its general solution is given by and . If it were -conjugate to its linearization then the leading order terms in the solution in the limit would be exponential but this is not the case. In three dimensions a vector field can even fail to be -conjugate to its linearization. An example of this is given in Hartman’s paper.

The proof of the theorem in the paper is actually not formulated in terms of ODE but instead uses a related ‘integrated’ version of the problem where the central object is a diffeomorphism with fixed point. The relation between these problems is explained in Hartman’s book ‘Ordinary Differential Equations’. If we start with a system of ODE we can consider its time one flow, which is a diffeomorphism. Suppose now that we can find a mapping which gives a conjugacy of the time one flows. Then it can be shown that it also gives a conjugacy of the entire flow and thus (under a differentiability assumption) of the vector field. Conversely a conjugacy of the vector field will give a conjugacy of the flow.

Here I want to look into Hartman’s theorem and its proof in some more detail. Consider a local diffeomorphism in two dimensions which fixes the origin. Suppose that it is hyperbolic with an eigenvalue of the linearization which is less than one and an eigenvalue which is greater than one. Using the stable manifold theorem for maps it can be assumed that the the stable and unstable manifolds of the fixed point coincide with the coordinate axes. Moreover, using the (integrated version of the) one-dimensional result already mentioned it can be assumed that the diffeomorphism is equal to its linearization on the axes. The mapping will be written in the following form. , . By what has already been said we can assume that and . This is the statement about the stable and unstable manifolds. The mapping used to do the transformation is of the form , and its inverse is of the form , . The condition for a conjugation implies a functional equation for the function . This gives rise to an iteration and the interation can be shown to converge. What is special about the case of the two-dimensional saddle? The argument with the stable and unstable manifolds works in any dimension. The fact that the vector field can be transformed to its linearization on these invariant manifolds uses the fact that the manifolds are one-dimensional. Othewise resonances could cause problems. In the paper the convergence proof is carried out in general dimensions under an additional assumption on the eigenvalues, (8.1) of the paper. These allow estimates to be obtained for the matrices making up the linearization. In the case of a two-dimensional saddle these matrices are scalars and the additional estimates are unnecessary.

There is a case where the non-existence of a conjugation of higher differentiability for a two-dimensional saddle can be shown by relatively elementary means and I want to show this here. Exercise 3.3.1 of the book ‘Nonlinear Oscillations, Dynamical Systems and Bifurcations of Vector Fields’ of Guckenheimer and Holmes asks to show that for a certain system there exists no smooth conjugation. Here I will do this exercise by showing that there is no -conjugation I have no idea if this level of differentiability is optimal in this example. The example is given by the equations , . The general solution is given by , . Suppose that there existed a conjugation of this vector field to its linearization at the origin defined by functions and . Then these functions satisfy certain functional equations. The equation for is . We are looking for a solution which is zero at the origin and whose linearization there is the identity. As a test take the first derivatives of these equations and evaluate the result at the origin. It turns out that the resulting equations are satisfied identically. Now do the corresponding procedure with second derivatives. The resulting equations imply that all the second derivatives of and vanish at the origin but no more than that. In looking at the corresponding procedure with the third derivatives the nonlinearity suggests that it could make sense to look at the derivative . In fact it can be concluded by doing so that , contradicting the assumptions. These calculations are justified if the transformation is assumed to be and this gives the desired result.

Conjugacy of one-dimensional vector fields near a zero

A vector field on the real line can be written in the form and in this sense it is equivalent to a function. From a geometrical point of view it is different from a function because it has a different transformation behaviour under diffeomorphisms. If a function is transformed using a diffeomorphism to get a function then . This is a pull-back operation, transforming a function on the range to one the domain. By contrast the transformation of a vector field goes in the opposite direction with the vector field written above being transformed to a vector field represented by a function with . If is the inverse of then this is equivalent to . Suppose now that has a zero at zero, i.e. . The zero is called non-degenerate if . This is an invariant statement since , where the last equality holds at a zero of . If we have two vector fields of this kind does there exist a local diffeomorphism defined in a neighbourhood of with which transforms one to the other, up to a multiplicative constant? This would follow if we could show that any vector field of this kind can be transformed to the vector field . The latter statement is not invariant but more convenient that the former for proving it analytically. It seems at first sight that this is a very simple question and I am surprised that although I spent many years working with differential geometry I cannot recall having asked myself this question before. It is the question of conjugacy of vector fields satisfying the given conditions.

The answer to this question of conjugacy is given in Proposition 2.9 of the book ‘Qualitative Theory of Plane Differential Systems’ by Dumortier, Llibre and Artes. This proposition concerns a differential equation of the form where and is a smooth function with . Solutions of this equation are precisely the integral curves of the vector field with the coordinate function and this is the general smooth vector field with a non-degenerate zero at zero. The statement of the proposition is that there exists a smooth function with for a smooth function with such that in terms of the equation becomes . This is the most general diffeomorphism whose derivative at is the identity. The condition on the derivative is of little importance since any linear transformation leaves the vector field invariant. The proof given in the book is as follows. If we transform the equation and impose the condition that the transformed equation is then we get the equation . This corresponds to the equation involving written above. This correspondence arises from the fact that the right hand side of a system of ordinary differential equations can naturally be thought of as defining a vector field. Let . Since is smooth and the function is also smooth. It follows that must be a solution of the equation . This has a unique solution with . This theorem is the special case of Sternberg’s theorem in one dimension, where there exist no resonances.

Definitions and politics

In mathematics definitions play a central role. Before starting a mathematical discussion it is necessary to fix necessary definitions. When mathematicians discuss other subjects they also often want to know the definitions of the terms being used and this is often difficult. It can already be difficult in physics and of course it is much more difficult in areas such as law or politics. A definition involves explaining something in terms of something else which is already known and already defined. In many areas this becomes difficult because so many terms used are poorly defined. The extreme case is that where terms are not defined at all but at just phrases connected to an emotion or a value judgement. In law or politics it is often hard for people to agree on a definition since different individuals may have concrete advantages or disadvantages depending on what definition is chosen. There is an increasing tendency for people to change definitions for their own benefit or use undefined terms as a weapon to discredit certain people or even to mark them as targets of violence.

A first example is redefining the meaning of the words ‘man’ and ‘woman’ so that they correspond to something else than they have done in the whole history of the languages they belong to. Since I live in Germany the example which is most familiar to me is the so-called ‘Selbstbestimmungsgesetz’. This is a German law which allows individuals to choose their own sex as it pertains to legal matters. This choice can also be changed as often as once per year. Changing the meaning of these central words immediately changes the meaning of many laws although the text of those laws remain identical. It offers many opportunities to a man who is prepared to make a corresponding declaration. Criminals who have done so (including sex offenders) have the opportunity to spend their prison time in a womens’ prison. It is possible to legally enter spaces reserved for women such as toilets and changing rooms. It is possible to compete in women’s sports. A second example is that political opinions (and people expressing those opinions) which do not belong to a narrow left-wing corridor are labelled ‘extreme right wing’.

A third example, and the one which motivated me to write this post, is the definition of ‘genocide’. What is the definition of this term, which was introduced by Raphael Lemkin in the early 1940’s? He was thinking primarily of the case of Armenia and Turkey. Later the Holocaust became the central example in people’s minds. Conflicts of interest such as those I have already mentioned have prevented the development of a clear definition which is widely accepted. What are the central elements of the concept? It has to do with the murder of a large number of people who are chosen on the basis of the fact that they belong to a certain group and not because of their actions. In a war between a country X and a country Y it may happen that large numbers of citizens of Y are killed by citizens of X because those people are fighting for Y. This does not fit the definition since the people were killed as a result of their own activity (fighting). The definition is also not fulfilled if large numbers of citizens of Y are killed just because they happened to be at the wrong place at the wrong time and there was no specific intention of the citizens of X to kill those people. The one aspect of ‘genocide’ which everybody agrees on is that it is very very bad. Thus it is an excellent label to use when trying to discredit people or mark them for attack. This is exactly how it is used by pro-Palestian activists as a way of attacking Israel. This is particularly absurd given that the modern state of Israel came into existence as a reaction to the Holocaust which is the clearest example of a genocide.

Normal forms of dynamical systems

Here I consider a system of ODE or equivalently a vector field on an open subset of Euclidean space. The question is in what way it can be simplified by a coordinate transformation. This leads to the theory of normal forms. My main source for what follows is the book ‘Qualitative Theory of Planar Differential Systems’ by Dumortier, Llibre and Artes. I will mainly concentrate on the two-dimensional case, as that book does. The aim is to simplify the vector field in the neighbourhood of a steady state, which is placed at the origin. Then we can write it in the form where is a matrix and vanishes faster than linearly at the origin. To avoid any subtleties with differentiability let us suppose that is smooth (). We want to apply a diffeomorphism to bring the vector field into the form where is in some sense as simple as possible. Let be the set of vector fields which are homogeneous of degree . The adjoint action of a linear vector field defines a linear mapping from to itself. We will be interested in this mapping in the case of the vector field defined by the matrix . In each degree choose a complement to the image of the linear map defined by the adjoint mapping. Then it turns out that there is a transformation which makes the sum of elements of the up to a certain order and an error term of higher order. It then follows using Borel’s theorem that there is a transformation which puts the whole -jet of at the origin into this form. The book also discusses a concrete choice for the complement .

We now restrict to the two-dimensional case. The analysis depends on the eigenvalues and eigenvectors of the matrix . Consider first the case that both eigenvalues have non-zero real part with the same sign. We can suppose without loss of generality that the real parts of the eigenvalues are positive (repelling case). Suppose now that is diagonalizable with real eigenvalues . It turns out that it is possible to reduce the system to the case where and . The normal form theorem shows that this can be done up to an error of infinite order. Sternberg’s theorem shows that it can be done exactly. Here can be chosen to be zero unless with . The case where is needed is the resonant case. Next suppose that is not diagonalizable. Then it turns out that there are no resonances and all non-linear terms can be removed. In other words the vector field can be transformed to its linearization by a diffeomorphism. The same is true in the case of complex eigenvalues with positive real parts.

In the case of a hyperbolic saddle the book starts by discussing the existence of stable and unstable manifolds, first in the analytic case and then in the smooth case. This can be used to rewrite the equation in a form where these manifolds have been mapped onto the coordinate axes. Only after that is the normal form theorem applied. In this case there can be resonances whenever the quotient of the eigenvalues is rational. There is a cautionary note that, in contrast to the cases previously considered, the theory of the smooth case cannot be straightforwardly be extended to the analytic case. This is related to the famous phenomenon of ‘small divisors’ in celestial mechanics. Returning to the smooth case note that when the quotient of the eigenvalues is irrational the system can be linearized. This completes the discussion of the hyperbolic case. I intend to return to the non-hyperbolic case in a later post.

A conflict of visions by Thomas Sowell

These days people often talk about how society is divided when it somes to various political issues. What is remarkable is a certain correlation. People who are on the same side in the case of one political issue often find themselves on the same side in the case of an apparently unrelated issue. This is a phenomenon I have thought about a lot and also discussed with other people. Recently I talked about it with Samuel Blotiu and he recommended me a book which is concerned with this subject and offerred to lend it to me. It is the book which is the title of this post. I was not immediately enthusiastic about reading the book but I was a little intrigued and I accepted his offer. I knew the name of Sowell only through some quotes I had seen on X but I usually liked them. When I started reading the book it immediately made a positive impression on me. The first sentence is ‘One of the curious things about political opinions is how often the same people line up on opposite sides of different issues’. Thus we meet exactly the theme I have just been talking about. I find Sowell’s style very pleasant. I also found myself confronted with unfamiliar and interesting ideas from the very beginning. When I started reading the book I stopped in a frame of mind which was similar to that when you stop eating chocolate. It tastes so good but you realise it makes sense to pause before eating more of it. I mentioned a correlation above and the aim of the book is to give some kind of causal explanation of it, even if that explanation cannot be complete. Here the notion of a ‘vision’ plays a key role and this was a new category for me. The short definition is ‘A vision is our sense of how the world works’. The author concentrates on two types of vision, the ‘constrained’ and the ‘unconstrained’. He emphasizes that it is not the purpose of the book to argue for one of these alternatives but instead to describe the differences between them. A first rough definition of this distinction is as follows. The constrained vision is that political action is limited by basic (negative) aspects of human nature. The unconstrained vision is that people can be made more altruistic by educating them, and that to an extent which is more or less unlimited.

The constrained vision emphasizes the value of tradition and ideas which have evolved within society. The unconstrained vision emphasizes the value of explicit reasoning. This might suggest that I, as a lover and proponent of rationality, should prefer the unconstrained vision. However that conclusion is too hasty. An important fact (one which has been emphasized by Hayek, someone who Sowell often quotes) is that when we make a decision we are often lacking much of the information necessary to make it in a purely rational way. Thus it makes sense to make use of other resources. There is a wisdom in society which has evolved. An example is the knowledge embodied in the development of the free market. If this kind of argument appears implausible it suffices to look at Darwinian evolution. In that context many things have evolved which are very beautiful and many which make use of remarkable mechanisms to attain goals. All this comes from the combination of mutation and selection. In a social context progress can come from the combination of chance developments and competition. The idea that those who know better should control the behaviour of others is limited by the fact that in many contexts those who know better still do not know enough in order to have a decisive advantage when making a decision. This does not conflict with the idea that in certain limited contexts experts can be expected to make better decisions. I would prefer to have my broken arm treated by a surgeon than by a carpenter. I now list a few other points from the first part of the book, which is a general discussion of the topic. Advocacy journalism is anathema to those with constrained vision since it is seen a a misuse of an entrusted role. Sincerity is so central to the unconstrained vision that it is not readily conceded to adversaries. The unconstrained vision tends to promote the interests of the young against those of the old.

The second part of the book is entitled ‘Applications’. The first chapter is concerned with the notion of equality. It starts with a discussion where I felt myself on very familiar ground. This discussion concerns the distinction between equality of process (which belongs to the constrained vision) and equality of result (which belongs to the unconstrained vision). The first of these was always what I regarded as the type of equality that I wanted to see in society. If I put this in mathematical terms, the first concept of equality is a property of the dynamics of the process while the second is an exercise is control theory. The chapter goes on to the idea that in the unconstrained vision the state of poorer people is seen as being influenced negatively by the activities of the rich. In the constrained vision the state of poorer people is influenced positively by processes in society to which some of the rich contribute decisively. The first variant is what I see being propagated by many German political parties, particularly those on the left. The second reminds me of statements of Rainer Zitelmann who has recently written a book related to this topic with the title ‘Zero
Sum Mindset’. (I have not read the book.) The second variant is also related to another theme which Zitelmann likes to talk about, namely envy as one of the worst diseases of society. Sowell explains the different attitudes of the two visions towards war and crime. In the constrained vision these negative phenomena lie in the nature of human beings and need no special explanation. Society should combat them directly, for instance by punishment of criminals. In the unconstrained vision these phenomena are caused by negative features of society and responsibility for crimes is shifted from the criminal to the society in which he was brought up and lives. In the constrained vision the rules embodied in laws have evolved from long experience and the judicial system should limit itself to implementing these rules. Attempting to implement perfect justice is not wise because it is a goal which is impossible to reach. Sowell strongly criticises the notion of social justice and quotes Hayek saying that it is a concept which ‘does not belong to the category of error but to that of nonsense’. In the book it is repeatedly emphasized that the difference between the constrained and unconstrained visions is not one of moral values but one of the belief in different types of causation of social and political phenomena. An example which occurs to me spontaneously is given by the following two statements. Hamas attacked Israel on the 7th of October because many Palestinians are driven by an evil ideology (constrained). Hamas attacked Israel on the 7th of October because of the way Israel had treated Gaza in previous years (unconstrained).

This is one of the most rewarding non-fiction books I have ever read. I already notice that when confronted with certain political events there is a resonance with ideas in the book. I expect that in the future I will have a better understanding of the actions of certain people, not in the sense of finding these actions more justified but in the sense of seeing how they fit into a wider pattern.

My closest encounter with a terrorist attack

Yesterday I heard a personal account of the aftermath of a terrorist attack from Sarah Levy. This made me think of an experience of my own. In 1995 I was living in Bures-sur-Yvette near Paris and working at the Institut des Hautes Etudes Scientifiques. I used to frequently travel into Paris in the evenings to go into the library of Pompidou centre. On one occasion the train I took stopped at the outer ring. It was announced that traffic was interrupted and that those who wanted to go into the centre of town should continue with the underground, which I did. No one seemed to know what was going on but from the conversations around me I got the idea that there had been a terrorist attack. I spent the evening in town and there was a very strange atmosphere. Nobody seemed to know what exactly was going on. There were red rescue helicopters landing on the square in front of Notre Dame.

What had happened is the following. A bomb exploded in a train of the RER line D, the one I usually took to go into town shortly before it reached the station Saint-Michel, the one I usually got out at. Eight people were killed and 171 injured. The helicopters were there to bring the injured to hospitals. If I had left home one hour earlier on that day I could have been in that train. The attack was carried out by a group called the Armed Islamic Group of Algeria. Around that time they also carried out several other bomb attacks in Paris and other parts of France. This was the time when suddenly dustbins became security risks and could no longer be used normally. Given the fact that there have been tens of thousands of violent attacks by people asserting their support for Islam it is not so surprising that I should come close to one of them. As far as I can remember in those days the attack was universally condemned in the media and the origin of the perpetrators was not kept secret so to avoid encouraging discrimination. Halcyon days.

The Parasitic Mind by Gad Saad

I have just read the book ‘The Parasitic Mind’ by Gad Saad. Its subject is political correctness and the decline of rationality and freedom of speech in the West. The author was born in Lebanon and escaped with his Jewish family to Canada as conditions in Lebanon became unacceptable for them. His parents made the mistake of returning to Lebanon once. They were kidnapped by terrorists and presumably tortured before they were able to return to Canada. Saad explains his strong commitment to freedom and to truth. In these things he is uncompromising and he was told by his mother that his commitment to truth was too strong. Here I feel sympathy for him since I also have a strong commitment to truth and have been criticised for it.

Having escaped from repressive conditions in Lebanon Saad was confronted in Canada with postmodernism in universities. He describes it as a parasitic mind virus. He compares it with physical parasites. One example is Toxoplasma gondii, which when it infects mice leads them to become friendly with cats, too friendly for their own good. I find the analogy quite appropriate. Our intellectual discourse is affected (infected) by a nefarious influence which could lead to the death of the West. As a young researcher who did not yet have tenure I considered whether I should concentrate my efforts on the US or the European job market. One reason I chose the latter was the spectre of political correctness in the US. In fact many examples presented by Saad indicate to me how right I was although I had no idea of the reality at that time. The events described in these examples are at first sight ridiculous but in the end really shocking.

I am not so fond of Saad’s style and choice of words which seem to me often a bit crude. For instance the word ‘lunatic’ occurs a bit too frequently for my taste. It may be that this is necessary for combatting political correctness since it is necessary to fight fire with fire. My own aesthetic feeling causes me to prefer the suave prose of Douglas Murray, as it is found in the book of his I reviewed recently. There are some formulations in Saad’s book which I liked, whereby I do not know which of those are his inventions. For instance there is ‘testicular fortitude’ and ‘tyranny of the minority’. In the context of political correctness he talks about self-flagellation. The idea of comparing the woke to flagellants is one which had previously occurred to me spontaneously.

One of the most interesting parts of the book for me is the seventh chapter. It contains a lot of interesting statistics. One simple question posed (and answered) is how many Jews there are in the world and how many Muslims. If someone had asked me that question without warning then I think I would have got the orders of magnitude right. I nevertheless found it striking to see the numbers. In a large part of the chapter data are presented which serve to illuminate the question about the nature of Islam ‘Is it a merciful, tolerant and peaceful religion or is it a religion of violence, intolerance and domination’. The numbers speak for themselves.

The last chapter in the book is an appeal for people to speak up. In the recent past I have increasingly tended to formulate my political opinions clearly in public without worrying too much about the consequences. In fact this has not had negative consequences for me. There are a couple of reasons for this. One of these is that my circle of contacts, physical and virtual, is quite narrow. The other is that as a professor of mathematics rather than, for instance, sociology I am far away from the main sites of infection in the universities. In fact speaking openly has mainly led me to interesting and pleasant contacts. Why did I read this book? I did so because what I had heard about Gad Saad had aroused my curiosity. Another alterior motive is that this was a kind of preparation for reading Saad’s upcoming book ‘Suicidal Empathy’.

Thoughts on Marcel Proust

As I mentioned in a post I wrote a long time ago the writer I admire most is Marcel Proust. This is something which has remained constant for very many years. In fact a long time ago he was part of a trinity in my personal literary heaven together with Virginia Woolf and James Joyce. At one time when I was rereading Ulysses I lived in Berlin and had a long commute to work. I tried reading Ulysses in the train but I sometimes had to laugh so much that I found it embarrassing. After that I decided to read the book only discretely and in private. Somehow in the course of time Joyce fell back a little in my estimation leaving the other two alone on the summit. In the end I feel that for me Proust is slightly ahead of Woolf. Some thoughts related to this can be found here. Special features of Proust’s work are that he only published one novel, ‘A la recherche du temps perdu’ and that that was extremely long. I have read the whole novel twice in the original and some parts of it more often. I recently started reading a little of it again and that is what has prompted this post. The question occurred to me whether Proust’s writing would still be as attractive for me or whether I might have changed so much over the years that that might have changed. The former is the case. As an example I quote a short passage from the first pages ‘comme ceux qui partent en voyage pour voir de leurs yeux une cité désirée et s’imaginent qu’on peut goûter dans une réalité le charme du songe’. [like those who go on a journey to see a city of their desire with their own eyes and imagine that it is possible to experience in reality the charm of a dream]. The author belongs to the group of people he is describing here and I do so too.

When reading ‘A la recherche du temps perdu’ you will not find much about mathematics. I seem to remember that Henri Poincaré puts in a cameo appearance, crossing the salon of Madame Verdurin, but I do not remember that anything significant was said about him. I do, however, see a connection between Proust and mathematics which is that it is often the case in the book that we experience Proust thinking in the way mathematicians do. I, as a mathematician, found myself feeling a sense of community in these cases. I cannot cite any examples. I am not trying to make an argument of literary criticism here – I am just describing my impressions. One idea which I see in Proust’s writing is that it often makes sense to identify things which are isomorphic. Proust presumably never encountered the word isomorphism let alone its formal definition but my impression is that he understood the meaning and significance of the concept very well. Another thing I want to mention is that in the last part of the novel, ‘Le temps retrouvé’ there is a scene which can be thought of as a spacetime diagramme. Not surprisingly it is a Newtonian spacetime, not a relativistic one. Proust was distantly related to Henri Bergson who also wrote about the concept of time. I read in the biography of Proust by Jean-Yves Tadié that Proust was irritated by the fact that people liked to compare him with Bergson. I have not read Bergson but I suspect that what he wrote about time was nothing but hot air so that Proust’s indignation was justified.

What is it that makes Proust’s writing so attractive for me? Both content and form are important. On the level of content I feel that I am very much on the same wavelength as the author in matters of philosophy. I really do not know to what extent I, in my youth, found my own ideas reflected in those of Proust and to what extent I simply adopted his ideas in forming my own. On the level of form I find his use of language exceptionally beautiful. It took quite some time for me to be able to absorb his long and complicated sentences. A favourite of mine (I claim no originality for my taste here – this is one of Proust’s most famous sentences) is the sentence which begins ‘Mais quand d’un passé ancien’. When I read the sentence it evokes a picture of a mountain stream, flowing rapidly downhill, turning repeatedly to avoid large stones until it flows into open water (l’édifice immense du souvenir). This image has nothing to do with the content of the sentence, it has to do with its rhythm.

Second Bonn conference on mathematical life sciences

I have just attended the conference mentioned in the title of this post. My general impression is that mathematics as I understand it is being more and more excluded by on the one hand masses of data and on the other hand by reliance on computers related to machine learning, AI etc. A cynical formulation would be to say that a research project is a machine for converting masses of data into complicated brightly coloured diagrams. The use of simple logical arguments to get real insights is becoming rarer. In my opintion this is not because such things are no longer possible or useful but decause they are no longer fashionable. Conversations with other older participants of the conference indicate to me that I am not alone in this opinion. Having got rid of some complaints let me now say something about some of the talks at the conference I liked best.

The first of these is a case where there were huge amounts of data involved and very complicated coloured pictures but there were also practical results which I found very impressive. The talk was by Bernd Bodenmiller from Zurich. In this work mass spectrometry techniques were used to produce very detailed pictures of the distribution of substances in slices of tumour tissue. I was surprised by one picture which showed the distribution of a conventional platinum-based chemotherapeutic agent within a tumour. While it was uniformly distributed through certain parts of the tumour it was more or less absent from others. Apparently cancer cells can develop methods to exclude this kind of drug from certain regions and thus survive. The one theme in the talk which caught my attention most was an application of this method to ovarial cancer. This is a very deadly cancer with frequent relapses after treatment. In the work reported on imaging techniques were use to distinguish different classes of patients and relate the differences between them to the rate of relapse. Beyond this predictions could be made which drugs might benefit which patients most. Patients were subjected to a kind of dual treatment strategy which would be illegal in Germany but which is fortunately legal in Switzerland. The idea is that on the one hand therapy decisions are considered in a conventional way and this information is given to the tumour board. Independently of this an analysis is done using the advanced imaging methods and this information also goes to the tumour board. These two sets of information are combined to make therapeutic decisions. In one case this method was applied to a patient already in palliative care, predicted to live for only a few more weeks. Five years later she is still alive and well. This is just one extreme case and the total sample size of patients is small. Nevertheless the preliminary conclusion is that this method leads to an large extension of the lifetime of the patients (I think a factor of four) in comparison to conventional approaches.

The second talk I want to mention is that of Becca Asquith. I had already heard a talk by her on a similar subject a couple of years ago and I wrote about it in a previous post. It has been observed that the KIRs an individual has can affect their ability to combat various infectious and autoimmune diseases, both positively and negatively, depending on the example. This is correlated to which MHC molecules the individual has. The subject of the talk was understanding the mechanisms behind these phenomena. One conclusion is that a determining factor is the typical lifetime of T cells. So how could KIRs modulate this lifetime? Two hypotheses are compared. One of these is that NK cells carrying the KIRs kill T cells, thus reducing their average lifetimes. Experiments were described which together with modelling, can decide between these two mechanisms. I find that this project was a beautiful combination of theory and experiment, exactly as I imagine such a project should ideally be. The whole thing, in particular the logical connections were very well described in the talk. At the end I asked the speaker why NK cells should kill T cells. Could this be of benefit to the organism or is it just a kind of collateral damage? My understanding of the answer, which I find plausible, is that any mechanism which can be used by the immune system to regulate its activity will be used.

The third talk was by Andreas Reichel, head of research at the company Bayer. He started off by mentioning a possible mechanism of action of a drug which is different to those commonly seen. This is to direct a certain protein to the proteasome so that it is destroyed. He then talked about the way in which candidate drugs are identified in the pre-clinical region. He mentioned a method in which a relatively simple ODE model can be used to obtain information. It can be used to find promising candidates. It can be used to suggest good doses for trials. (Sometimes increasing the dose produces no effect of the kind desired.) It can be used to choose optimal times for taking blood samples when testing candidates. Apparently it has been possible to convince decision makers that this theoretical work is something they can really profit from. For me this is a good example of how (relatively simple) mathematics can be used to make a significant contribution to a practical task such as drug discovery. If someone wants to develop models of this kind or apply them in an intelligent way then they need to things from analysis which I teach students on a day to day basis.