XII. Why does Toxoplasma shift personality traits in opposite directions in men and women
From the very beginning of our Toxo human behavior studies, we were puzzled by the opposite shift in psychological factors in men and women. Unfortunately, the reviewers of our manuscripts were very puzzled as well; and so this baffling observation often cost us a possible publication. Neither editors nor reviewers are fond of unexplained phenomena. It’s fortunate that there are a great many scientific journals; so if an author is patient enough, he will eventually get his manuscript published (see Box 39 How I outwitted unwelcoming editors and reviewers).
One of our hypotheses to explain this phenomenon was based on the idea that Toxoplasma causes the same personality changes in men and women, but that one of the genders lies about these changes. They may be aware of their new, Toxoplasma-induced personality traits and be unhappy with them, or consider them to be inappropriate; so when filling out the questionnaire, they try to mask these changes. Unintentionally, they overdo their facade and the changes show up as the opposite of what they really are. For this reason, we were interested in monitoring these psychical changes in a more direct way; for example, by observing a person’s behavior instead of determining it through a psychological questionnaire.
One experiment was practically begging to be enacted; it was also one of the first which we conducted on the students. We set a time to meet with the students (for testing or for taking a blood sample) and observed how promptly they arrived. We expected that people with a lower Superego Strength (Cattell’s factor G) would arrive later, whereas those with a higher Superego Strength would arrive mostly on time, or even early. The results of the questionnaire survey had indicated that men had a weaker and women had a stronger superego. If one of the genders was lying, and Toxoplasma actually did have the same effect on the Superego Strength of men and women, then infection should correlate with punctuality in the same way in both genders. In other words, both Toxo positive men and women should arrive more punctually or less punctually than Toxo negative men and women. The results of the behavioral experiment, however, corresponded with those of the questionnaire survey. Toxoplasma seemed to have an opposite effect on the Superego Strength of men and women – men appeared to be less punctual, and women more.
Our other experiments were more sophisticated and combined direct behavioral observation with questionnaire studies. When testing Superego Strength, our questionnaire quizzed the students on their knowledge of various social norms. The logical assumption was that only people who know social norms can follow them. Students had to answer various questions, such as who should first enter a restaurant, a man or a woman; who should be first in going down the stairs; how you arrange the silverware for dessert; how you hold a spoon and fork when using both to eat it; whether or not you should always remove your gloves when shaking hand, etc. In addition, we observed how long they were willing to spend on our questionnaire. We expected that people with a weaker superego would skip through a questionnaire that forced them to remember all sorts of details (see below), and end up finishing more quickly than people with a stronger superego, who are more conscientious. Furthermore, we observed how much care each person had given to their appearance. We wrote down the observations in a chart: how new and clean were the pants, skirt, sweater or shirt they were wearing; how well-kempt the person was; whether or not they wore accessories – rings, necklaces, brooches, etc. Without informing the test subjects ahead of time, we observed the expensiveness and cleanliness of their clothing, and finally the overall tidiness of a person’s appearance (Box 59 A blind and a double-blind experiment).
When testing sociability (warmth, extroversion, i.e. the analog of the Cattell’s factor Affectothymia), which the questionnaire determined to be lower in infected men and
Box 59 A blind and a double-blind experiment
It’s clear that we couldn’t inform the people we were observing for punctuality and appearance that we were doing so. Even knowing if they were Toxo positive or negative, could influence their behavior. Of course, under no circumstance could we let the students observing the test subjects knew which of the people was Toxo positive or negative. Judging the costliness and cleanliness of clothing already isn’t a very objective matter, and if the observer knew which people were infected, he might also have a previous conception of their behavior. Therefore, the whole study had to be conducted as a double-blind experiment. We told the test subject whether they were Toxo positive or negative only after the study; and during the study, only one of the workers knew who was Toxo positive and negative – and this worker wasn’t involved in observing the subjects. A double-blind experiment, in which neither the test subjects, nor those giving the test, know who is in the experiment and who in the control group, should be applied whenever possible. For example, when testing a new drug, it’s clear than the test subject can’t know who got the drug and who got the placebo, a pill which doesn’t possess the active ingredients. Just the belief that one is being given medicine often has a positive effect on a person’s health, even when the drug itself is ineffective. For this reason, people in the control group must get pills that look the same, but don’t contain the tested drug. A number of studies also show that the doctor himself can influence the test subjects, if he knows which people are getting the drug versus the placebo. Hence an experiment must be set up so that nobody directly involved in the testing knows who is given what.
higher in infected women, we asked questions like: how many parties you’ve been to lately; how many text messages you’ve sent over the past days; how much money you spend each week on phone calls. Then we tested whether they remember the names of various actors (we projected that actor’s last name on the computer screen and had the test subject supply the first name). We ascertained whether they remembered the birthdays and name days of their close friends and family, and so on.
When measuring Suspiciousness (Vigilance), which the questionnaire had found to be higher in infected men and lower in infected women, we used a number of simple experiments, which were quite entertaining to come up with. For example, we asked the person to taste a liquid in a test tube, and observed whether or not the person would do so, or at least ask what was in the test tube or why he should drink it. Or we asked the test subject to sign a blank piece of paper, and then looked to see whether or not they complied; and if they did, whether they signed in an upper corner of the paper, so that nothing could be written above their signature, or if ingeniously wrote their name in the center. To evade possible prosecution, we showed the person that the paper was torn up just after they signed it, so that they might not live out the rest of their lives worrying that we could somehow take advantage of their signature. Furthermore, we asked test subjects if we could photograph them, and noted how willing they were to agree. Similarly, we observed how willing they were to undergo anthropometric measurements. For the most complicated experiment, which we used to determine trustfulness, we had the test subject sit before a dusty ancient machine. We carefully set the dial to display 3000 volts, and asked the person to hold an enormous electrode in each hand, and, “when they were ready,” to press the button on one of the electrodes. The electrodes, of course, weren’t connected to a current. Even we were surprised at how many of our students followed our instructions without comment or complaint, and, hardly hesitating, pressed the red button.
From the above questions and observations, we created factors analogous to those measured by Cattell’s questionnaire. We measured a person’s Conscientiousness, which should correspond to Superego Strength (Rule-consciousness, G); as well as an analog of Suspiciousness (Vigilance, L), and then an analog of Affectothymia (Warmth, A). These three were the Cattell’s factors which most obviously shifted in opposite directions in men and women. Our results were unambiguous, and almost exactly matched the results of the earlier psychological questionnaire. Infected women were more open, sociable and warm-hearted in comparison with uninfected women, while infected men were more introverted and reserved than uninfected men. The results were similar in the case of the Rule-Consciousness (Super Ego Strength). Here the differences were particularly obvious when observing how the test subject dressed. Toxo positive women came to the testing very tidied up, wearing newer and more expensive clothes, whereas Toxo positive men arrived in ripped up jeans and were less well-groomed – in comparison with both male and female uninfected peers (Fig. 30) (51). In the case of Suspiciousness (Vigilance), the results were more complicated: the effect of toxoplasmosis on men and women depended on whether the person came from a village or a town
Fig. 30 The effect of toxoplasmosis on the behavior of students; specifically, on how much care they put into their appearance. The cleanliness and neatness of dress was estimated on a 7-point scale by female graduate student. She had no knowledge about the infection status of the test subjects; not even the test subjects knew this at the time. The graph demonstrates that infected female students were more neatly dressed than uninfected females. In contrast, infected male students took less care about their appearance than uninfected males. The difference in Toxoplasma’s effect on each gender was statistically significant; the effect of toxoplasmosis was stronger in male than in female students.
Based on our results, we came to the conclusion that the differences we found between men and women, in regards to the effect of toxoplasmosis on the human psyche, are not due to one of the genders lying in the questionnaire. It appears that toxoplasmosis has a different effect on the actual behavior of men versus women.
Nevertheless, we weren’t satisfied with these results, and tried to verify our surprising discovery that the two genders are often oppositely affected by latent toxoplasmosis, using the method of experimental games (52). Experimental games are fairly recent, and used primarily in experimental economy, though can have other uses. Unfortunately, it later became apparent that they are but rarely used in psychology and ethology, so publishing our unusual results obtained through such an unusual method proved to be a near superhuman effort. At first, our study used two games – Dictator and Trust games.or the Dictator game, we set 12 students in front of 12 computers so that the students couldn’t see each others. We even had the first six students come fifteen minutes early, so we could have them sit at computer tables placed behind a screen which divided the room in half. The remaining six students sat on the other side of the screen, opposite the first six. A student on one side of the screen always played a student on the opposite side of the screen; in this way, each player couldn’t see who his opponent was. Furthermore, people on the same side of the screen were given privacy by testing dividers, so that neighbors couldn’t see each other’s computer screens.
The students were told that the computer would divide them into pairs of “donor and recipient,” that they’d participate in several rounds and be paired up with a different opponent each round. Each round, the proposer was given “an endowment” of 50 cents and was free to give how much he wished to give to the recipient. Nice donors would divide the sum equally with the recipients, but the mean ones often kept every cent. The roles of donors and recipients switched in every round, but the pairs were different. Of course, it was real money, and the students were allowed to keep it at the end of the game as a reward for their participation.
As often happens in this sort of game, it started out with the donors giving the recipients a fairly large amount of the money, but as the game progressed, they grew less and less generous. In the last round, most of the donors gave their recipients nothing, or the smallest sum possible, 5 cents (though this was probably more out of spite than any sort of sympathy). There were 12 rounds, so each player was a donor six times and a receiver six times. We expected that the Toxo positive woman donors would give their recipients more money, and that Toxo positive man-donors would give their recipients less money, than the Toxo negative men and women.
The second game, the game of Trust, was a bit more complicated. The basic set-up was the same, but this time the computer formed pairs of proposer and responder. The game began with the proposer getting 50 cents, and he could choose how much of this money to give to his partner. The money given to the responder was tripled by the computer, and then the responder could choose how much of his money to give back to the proposer he was paired with. Let’s say that the proposer gave the responder all 50 cents. The responder would then end up with $1.50, and could choose how much (if any) to give back to the proposer. There were also 12 rounds, and the players were paired up differently each time. In the game of Trust we observed how much the proposer invested, and with how much of the tripled money the responder would reward the proposer. Once again, we expected that Toxo positive men would give away less of the money and that Toxo positive women would give away more of the money than the Toxo negative controls.
At first, the results of both games were a bit surprising. It turned out that in the Dictator game, toxoplasmosis lowered the sum of money which both man- and woman-donors gave to their recipients. In contrast, in Trust, infected woman-responders gave the proposers more money, whereas infected man-responders gave away less of their money. There was no difference between the amount of money that Toxo positive and negative proposers gave to the responders. In the end, we managed to formulate a reasonable explanation for these results – but little wonder, since we scientists are well-trained to find logical explanations for unexpected results. Our hypothesis is that latent toxoplasmosis introduces a mild, but constant stress factor. It’s known that stress causes men decrease their social interaction, whereas stressed woman seek support in their social surroundings, becoming more sociable. During the Dictator game, social interaction didn’t play a great role in the relationship between donor and recipient. Players got to alternate between the part of donor and recipient, so there is no reason to think that highly sociable people would have given their recipients more money than unsociable people. On the other hand, social interaction played a substantial role in how much money the responder gave back to the proposer in the game of Trust. By investing his money, the proposer gives the responder his trust, and so one might expect that more sociable responders would be less liable to break this trust. So if our stress hypothesis is true, Toxo positive men and women playing Trust should have a tendency to give proposers less money. The stress hypothesis, proposed by Jitka Lindová, corresponds with our results. This of course, is unsurprising, since the hypothesis was suggested after the study, in an attempt to explain the unexpected results. In our defense, we have followed up with unrelated experiments to test the stress hypothesis, by measuring the levels of stress hormones and by using our favorite questionnaire method. That, however, is for another chapter.
Later we used experimental games to observe the effect of toxoplasmosis on the altruism of a student. Altruism is behavior which helps one’s surroundings at the cost of the individual. Altruism has been observed not only in humans, but in a number of other species. The advantage of some forms of altruism was obvious, but other forms long resisted explanation. For example, reciprocal altruism is clearly in the interest of the individual. When North American vampire bats (of the genus Desmodus, which subsist from the blood of mammals or birds) return from a uccessful hunt, they regurgitate some of the blood for the less fortunate members of the bat colony. The next time these bats have an unlucky hunt and return hungry, the successful members of the colony will help them out in return. It’s not true altruism, because bats can distinguish between each member of their colony, so a bat that didn’t share what he acquired from successful hunts wouldn’t get any help from other bats. Three consecutive unlucky hunts are enough for a bat to starve to death. It’s also easy to explain altruism in regards to offspring, siblings or other relatives. When the genes for altruism cause the individual to help his relatives, which most likely also possess copies of these genes, it increase their chances of spreading to the next generation. So this also is not true altruism, because each gene has “calculated” (or rather natural selection has found) that it pays off when its individual sacrifices a little for the survival of other copies of the gene in that individual’s relatives.
But in some cases an organism helps individuals that are not his relatives, and which can’t be expected to someday return the favor. Evolutionary biologists have long known that the existence of this form of altruism is an evolutionary mystery. It’s clear that a population with many such altruists is better off than one without them; the first type of population flourishes at the expense of those with few altruists. But it’s also clear that selfish individuals have an advantage in the first type of population, for they are helped by the many altruists without having to compromise their own biological fitness in order to help others. In a population with altruists, the selfish individuals should reproduce the most and eventually push out the altruists. Over time, theoretical biologists have shown that there exist conditions under which natural selection for the benefit of the community prevails over selection for the welfare of the individual. Under such conditions, genes for altruism can prevail (Box 60 When does altruism prevail over selfishness?).
Box 60 When does altruism prevail over selfishness?
Altruists can triumph over selfish individuals primarily when the species consists of a large numbers of small populations, which constantly pass in and out of existence. The new populations must form from small groups of individuals which come from other populations. Ideally, the new populations form from individual migrants which left their original population. Populations with a lot of altruists prosper, and produce more migrants than populations which wither due to a lack of altruists. The populations dying out due to a lack of altruists produce fewer migrants that could start up their own population or join another population – in other words, highly selfish populations are less likely to produce migrants that could spread the selfish gene. When a selfish individual appears in a population of altruists, he begins to reproduce rapidly. But as his offspring start to spread throughout the population, the population will deteriorate. There is a less of chance of altruists existing in a species made up of just a few populations, with a new population forming only when an old populations splits about evenly in two.
For a long time, it was thought that conditions under which altruists could proliferate were rather rare. Recently, a small but international, interdisciplinary scientific team (made up of myself and Tomáš Kulich of Comenius University in Bratislava, Slovakia), found that in sexually reproducing species the conditions are more advantageous for the spread of altruists than it seems at first glance (53). Our model comes from the theory of frozen plasticity (see Box 35 How I refuted Darwin) and presumes that a particular trait – in this case, altruism – depends on multiple genes. Previous models were set up rather unrealistically, as though altruism depended on only one gene. Assuming that altruism depends on several genes, a family of altruists may give birth to a selfish individual, and vice versa. In this scenario, the survival of altruists isn’t constantly penalized – if altruists gave birth to only altruists, their offspring would survive and reproduce less than the selfish individuals, and their genes would eventually die out. Altruists appear in a population seemingly randomly, at a frequency which maintains a constant number of altruists in that particular population. The populations of a species can compete amongst each other, as to which population has more altruistic genes; while the individuals within a population don’t need to compete based on who is more selfish and thus more likely to reproduce, because neither altruism nor selfishness is particularly heritable.
Experimental games are useful for studying altruism – we, for example, used two such games, called Public goods and Public goods with punishment. Public goods is a simpler game. 12 students sit at separate computers so that they can’t see each other. Each person is given a dollar, and is free to contribute how much he wishes to the public pool. The rest he keeps. The total sum of these “public goods” is doubled and equally distributed among the 12 players, regardless of how much each person contributed. It is most advantageous for the group when everyone contributes all their money to the public pool – if each person contributes their dollar, then everyone will get 2 dollars back. But for the individual player, it is most advantageous when everyone but him contributes all their money. The 11 dollars in the group pool will be doubled to make 22 dollars. At the end of the round each person will get a twelfth of the 22 dollars, a total of about $1.83, except for the person who didn’t contribute. He’ll get $1.83, plus the 1 dollar which he kept – a tidy sum of $2.83. Our students behaved rather typically. In the first round they put a fairly large amount in the group pool, but very quickly they realized that altruism – and primarily, a belief in the altruism of others – isn’t very advantageous in today’s world. So the average contribution to the group pool plummeted over the course of six rounds. The amount contributed in the Public goods game reflects each player’s belief in the altruism of others, rather than the extent of their own altruism. On the other hand, Public goods game with punishment better reveals the altruism of the individual. This game begins the same as Public goods game. But after the players finish contributing to the group pool, each person’s contribution is made public (of course their real name isn’t made public; rather, they are identified by numbers, which change every round). And now things get interesting. Players can buy the right to punish another player – which means that the person who takes it upon himself to punish someone sacrifices a substantial amount of money. So if one player skimps on his donation to the public pool, he may pay for it twice over; if he is punished then he loses all the money which he withheld in that round. Giving the punishment is altruistic act. The person doling out the punishment must pay for his right to punish, and the possible benefit that comes from disciplining the penny-pincher applies equally to the rest of the players (which didn’t have to pay for it). Once again, our students behaved rather typically. They frequently, altruistically punished selfish players. As result, people contributed more money to the public pool than in the simpler game Public goods (without punishment), and their contribution remained consistent throughout.
So what did our experiments show? What is the effect of latent toxoplasmosis on the altruism of our students? Once more we saw differences between men and women. In the game Public goods, which threatened no punishment, toxoplasmosis had no effect on the amount of money put by men or women in the public pool. But in the game with punishment, Toxo positive men contributed significantly more in the 2nd through 4th rounds, and punished more in the 1st through 3rd rounds than did Toxo negative men. Toxo positive women, on the other hand, contributed less and punished more than did Toxo negative women in all the rounds – the differences, however, were not significant. As could be expected, contributions to the public pool decreased over the course of Public goods (men went from 70 cents in the first round to 20 in the last, whereas women went from 60 cents to 15 on average). In contrast, contributions stayed fairly high in the game with punishment available (the contributions of men stayed around 70 cents, whereas those of women even increased from 55 to 65).
We used these two games over the course of about five years. For one, the students liked them. If you’re going participate in a test, why not earn a couple of dollars? These games helped us attract more students to our experiments. But the main reason we continued using these tests, was that the results led us to a number of interesting phenomena, quite unrelated to toxoplasmosis. These included Justine’s effect and the existence of perverse punishers. Perverse punishers among our undergrad students? Great heavens! (Box 61 How we discovered perverse punishers among our students.)
Box 61 How we discovered perverse punishers among our students
During the experimental games, it was important that we always had 12 players. Several times it happened that one of the students didn’t show up, and even that the alternate didn’t come. When this happened, we had to use one of our people to act as the missing student; he had to follow a scenario which we had prepared ahead of time. The scenario told him how to behave in each round (how much money to give to his partner, or put in the public pool). One time our head statistician Aleš Kuběna acted as the missing student, and he noticed an interesting phenomenon: twice he was punished when had contributed the maximum amount of money, 1 dollar, to the public pool. After further examination, we verified that this wasn’t a chance occurrence, but happened in almost every game. In a group of 12 students, there were almost always two or three perverse punishers who were stupid or vindictive enough, that they didn’t hesitate to use their money to punish the most altruistic of the players. When we studied up on this phenomenon, we soon found that it is not specific to the students of the Prague College of Natural Sciences. It was noted, though not particularly elaborated on, by other researchers as well. If we consider the popularity of common saying, “no good deed goes unpunished,” we can see that people have long understood this phenomenon’s existence. Before we could finish, evaluate and publish our study, we found that other studies, which focused particularly on the existence of perverse punishers, had been published in the meantime. One of these studies showed that the prevalence of perverse punishers correlated negatively with the development of the civil society in a given country. It reported that the fewest perverse punishers occurred in Scandinavia, whereas the most were found in Post-Soviet Republics. Thankfully, the Czech Republic was not represented in the study. Unfortunately, most of these published studies used about four players in their experimental games, so it was difficult to determine who was the primary target of the perverse punisher. Our experiments involved 12 players, so we could organize the results of each round according to how much each player contributed to the public pool, and then look at who was most likely to be the target of punishment. We found, of course, that the people who contributed least to the public pool were most likely to be punished. The more that a player contributed to the pool, the smaller his chance of being punished. People least likely to be punished contributed about the fifth largest sum of money (out of 12 people). But players that contributed more than this started to have an increased probability of being punished – so the greatest altruists, that contributed the most to the bank, had a relatively high risk of being targeted. We named this phenomenon Justine’s effect,after the altruistic heroine of Marquis de Sade’s novel Justine, who is the victim of constant attacks of fate and her surroundings. The effect describes this: the further a person strays from the expected, or slightly above average, level of altruism – the closer they get towards maximum altruism, the more likely they are to summon the wrath of a perverse punisher. I think that the author of the above-mentioned saying about the punishment of good deeds, as well as the fictional Justine (had she not been struck down by lightning at the end of the book) would be touched by our findings.