The Errors of Scientists

This essay delves into the pervasive presence of errors in the annals of scientific history, spanning from its infancy to contemporary times.

By Roberto Assagioli, (Docs. #23323 and 22324 Assagioli Archive – Florence). Original Title: Gli Errori Degli Scienziati [i] Published in Psyche – Journal of Psychological Studies -Year II. – No. 4 – July-August 1913. Translated and Edited With Notes by Jan Kuniholm[ii]

Abstract: This essay delves into the pervasive presence of errors in the annals of scientific history, spanning from its infancy to contemporary times. It underscores the significance of studying these errors, particularly for psychologists, in elucidating the intricacies of scientific thought processes, belief formation, and the psychology of human mentality. The author highlights the potential utility of such studies in advancing scientific progress by uncovering overlooked errors and cautioning against future ones. While acknowledging the complexity of categorizing errors, the essay provides a glimpse into various classifications proposed by philosophers like Bacon and Locke, emphasizing the multifaceted nature of scientific fallacies. From errors of observation to misinterpretation of data and logical fallacies, the essay delineates an array of common pitfalls scientists encounter. Additionally, it discusses errors stemming from linguistic ambiguities and emotional biases, alongside professional errors inherent to specific disciplines. Despite the prevalence of errors, the author cautions against a blanket skepticism towards science, urging scientists to remain vigilant against potential sources of error. Ultimately, this essay suggests that while errors can impede progress, they can also serve as catalysts for discovery and stimulate critical inquiry within the scientific community. It also is evidence that Dr. Assagioli was himself aware of the need for and use of clear thought and appropriate conceptions and practice, which some of his critics thought missing from his own work.

Anyone who thinks about the history of science cannot fail to be struck by the large number of errors that have been made by scientists. And not only in the distant past, when science was in its infancy, when instruments of precision were lacking and the scientific method had not yet been well formulated and defined; but also recently, even in our own day.

The study of such errors is of great and wide interest, especially to psychologists. It provides numerous data on the nature and procedures of scientific activity. It throws much light on the formation of beliefs and convictions, on the disturbing influence of attitudes and affective states over knowledge, on the psychology of mentalities, etc. It can then prove most useful to the progress of science, both by helping to uncover errors that were previously missed, and by warning against possible future errors.

A full treatment of this subject would require a large volume; but while waiting for some well-meaning scholar to prepare it, I do not think it useless — also as an inducement to that work — to recall some of the most important species of scientific errors and to cite a few examples.

Philosophers and scientists have often written about error, but in general they have tried more to establish what it is and what relationship it has to truth; that is, they have studied it mainly from the epistemological point of view. There is indeed no shortage of analyses and classifications of errors, but these mostly concern only errors of logic.

Among the more general classifications, I will mention those of Bacon[iii] and Locke.[iv] The former distinguishes errors into four classes: the idola tribus [“idols of the tribe”], which are illusions arising from the very constitution of the human spirit that are common to all people; the idola specus [“idols of the cave”], or illusions arising from individual temperament, education, and the various dispositions and states of mind of each person; the idola fori [“idols of the marketplace”], which are dependent on transmission by language; and finally the idola theatri [“idols of the theater”], which are due to the academic and systemic spirit.[v]

Locke too finds four causes of error, from another point of view: 1. The lack of proofs; 2. The lack of ability to use them; 3. The lack of willingness to use them; and 4. The miscalculation of their probabilities. This last class is in turn divided by Locke into four categories: a) Doubtful propositions assumed as principles; b) Accepted hypotheses; c) Ruling passions; and d) Principle [accepted] on authority.[vi]

These classifications are not entirely satisfactory, for various reasons on which I will not dwell, for the sake of brevity, and they are also very incomplete. Indeed, a complete and rational classification of errors is very difficult, both because it presupposes an accurate analysis of all their many varieties, and because it can be made with very different criteria and from very different points of view. Indeed, I believe that a single classification is insufficient and that at least two are needed: a “descriptive” and a “genetic” one.

I certainly do not intend to give such classifications here; I propose only to enumerate, without any claim to methodological rigor or completeness, the main varieties of errors that scientists are most often subject to.

There are, first of all, errors of observation, which arise either from the imperfection of the instruments used, or — more often — on the imperfection of the observer’s sense organs, or on his sensory illusions.

Then there are numerous errors in the interpretation of the observed data. To this category belongs, for example, the error, committed not infrequently by histologists,[vii] of interpreting certain characteristics observed by them in microscopic preparations as belonging to the living substance, whereas [in fact] they arise from post-mortem alterations or on the manipulations to which the substance has been subjected.

Errors can creep into even the most seemingly rigorous and “objective” experiments. I will mention, for example, what Ochorowicz[viii] in a recent paper called “experimental tautologies.”[ix] These are experiments that are unconsciously influenced by the theoretical preconceptions of the experimenter, so that the results are always those expected by him and are mistakenly believed by him to be “evidence.” An example of this occurred in some of the experiments that Charcot[x] did on hysterics. [The evidence he found arose from phenomena that] he himself created directly, or indirectly by means of the general suggestion of the environment, [and produced] the states and symptoms he expected.

Another category, which is very broad, is that of errors of logic. These “paralogisms” are mostly formally the same as fallacies, differing from them only because they are unintentional. Numerous species of such errors of reasoning are described in treatises on logic; I will mention only a few.

There are first of all arbitrary generalizations (corresponding to the “fallacy of induction” or “fallacy of imperfect enumeration” of logicians).[xi] On the basis of an insufficient number of observations or experiments, often disguised under convenient “percentages,” a principle is affirmed, a general law is formulated . . . until other facts come to light, which show the non-existence or relativity of the [purported] principle or law. This kind of error is very frequent, and it would be easy enough not to commit it in its gross forms if more prudence and more critical sense were used; but it is impossible to avoid it altogether, since it has its root in the very nature of induction, which is the firmest pillar of the scientific method.

Closely related to the preceding are errors due to misuse of analogy. Errors due to excessive anthropomorphism fall into this category. There has been much outcry against such errors, especially by “positivists;” but not infrequently they have seen the twig in the eye of others without noticing the beam in their own. Analogical inference is often very similar to induction, so much so that in many cases it is difficult to distinguish one from the other. Thus it has often been believed that the inductive method, the only one that is considered to be a “truly scientific” method was being used by someone, when in fact the person was using the despised “analogy” method without realizing it. A shrewd and prudent use of this [i.e. analogy], as shown, among others, by Adolf Wagner in his study on “Psychobiology as Science”[xii] is legitimate[xiii] — indeed it is necessary in science. But to avoid its abuse, one must first of all realize that one is using it.

Some abuses of analogy may be partly considered as errors due to excessive imagination. Many philosophers, including, with particular insistence, Malebranche,[xiv] have warned against these errors.[xv] Nevertheless, philosophers and scientists continue to make them, but this should not surprise us; not only because, as is well known, men are wont to profit little from good warnings, but also because — as many authors have shown[xvi] — bare facts do not yet constitute scientific knowledge, and a certain use of the “scientific imagination” is necessary to coordinate and truly understand them. Keeping this within the right boundaries is not easy, and therefore in scientific works one frequently encounters errors due to the too unbridled fantastic activity of their authors; even of those who most care to be positive. For example, does it not require at least daring flights of imagination . . . to conceive, as Loeb[xvii] did, even the mere possibility of reducing human higher psychic facts to “tropisms?”[xviii] While there is still lively discussion of the brain localizations, do not those scientists show an imagination that novelists would envy as they construct true histological mythologies, imagining an anatomical seat of the highest and most complex psychic states and processes — even . . . asserting that the subconscious is located in some hypothetical “polygon” or layer of brain cells?

Another group of errors, much akin to that of arbitrary generalizations, indeed sometimes mixed with it, is that of simplistic theories, which are so numerous in science. They are built on the false assumption that the data and elements that are known or taken into consideration at a certain time constitute all actually existing and possible data.[xix] This, too, is a mistake that could often be avoided if scientists had a more developed a sense of the complexity of issues and had less self-confidence, remembering how insufficient human knowledge is in the face of the infinite richness of reality.

I will not cite many particular examples of such erroneous “simplistic theories;” for who does not know of them? I will only recall, as a psychologist, how many times mistakes have been made, due to not taking the psychological, human element into account. This is what historians, economists, and sociologists who are inspired by the principles of historical materialism have done and cheerfully continue to do.[xx]

Some of the errors of the latter group and many others, which are very frequent in science, can be seen as arising from the erroneous or indiscriminate use of the concept of causality.

The problem of causality is one of the most serious and debated philosophical problems, even recently.  But in spite of this many scientists do not understand it, and make every herb a bundle [i.e. lump everything together] with a mind-boggling nonchalance. It is mainly doctors who hand out the “license” of cause with the same generosity with which government ministers award knight’s crosses — they call every circumstance even remotely connected with a given fact a “cause.” The “cause” of tuberculosis is as much Koch’s bacillus as alcoholism, as much organic predisposition as a poorly ventilated environment, as much humidity as exhaustion. It results — and I could cite examples — from taking the effect . . . as the cause.[xxi]

Such erroneous or indiscriminate uses of the concept of causality can be divided into several groups. There is first of all the error of evaluation or perspective, just now mentioned, which consists in calling the “efficient causes” and the “concomitant causes,” the immediate and the remote causes, the “conditions” and the “occasions” [all] by the same name and putting them, as it were, “on the same plane.” Then there is the error designated by the well-known phrase post hoc ergo propter hoc,[xxii] that is, that of taking as a cause a circumstance which preceded an effect but which in fact exercised no determining action on it. Finally, I will recall what might be called “stopping at the immediate (or proximate) antecedent;” such a fallacy occurs when one stops at the first one or few links in a causal chain composed of many links, and ascribes to them the significance that the more remote ones should have instead. Many of the flood of pseudo-explanations of which scientists are so full of are due to this fallacy. However, it is impossible to escape completely even from this fallacy, since our knowledge of causal chains is always necessarily incomplete.

This confusion about the concept of causality depends not only on the lack of clear ideas, but also on the imperfection of terminology. This leads us to recall the numerous errors due to language. They, which Bacon calls idola omnium molestissima,[xxiii] have been dealt with by several ancient and modern philosophers; although I believe that scientists still do not sufficiently realize how imperfect and insidious language is as an instrument of expression, when it is not used for practical or artistic purposes, but is required to express knowledge and ideas exactly and rigorously.

Among the very numerous varieties of verbal fallacies I will mention only those most often found in scientific works. In first place are the errors due to ambiguity of terms. It can be said that almost any word which does not designate a concrete fact can be taken in different senses. Sometimes this is self-evident, as in the case of the word cause (yet even this is not enough to avoid errors!), but in many cases it is not. For example, take those famous differences, often subtle and ill-grasped and yet of paramount importance, between the “strict sense” and the “broad sense” of the same word, which are a continual source of confusions, misunderstandings, and errors. Rather than the adoption of an “official terminology” drawn up by scientific commissions (which would not be without its drawbacks and very difficult to implement, as appears from the failure of the practical attempts made so far), the custom of indicating exactly the sense they intend to give to the main terms they use should be more widespread among scientists and would be worthwhile to remedy these errors.

Another variety of fallacies in which language is largely to blame are verbalistic pseudo-explanations, so detrimental to the progress of science, because they give a false security. For some scientists certain words are real magic wands that open every door, that solve every question, that “explain” everything. Recall for example the word “suggestion” — what has not been “explained” with it! — hypnotism, hysteria, mediumistic phenomena, psychotherapy, lying, demonstrations of thinking horses . . .

There is another category of “explanations” whose insufficiency or erroneousness is increasingly being recognized: illegitimate reductions of facts and “forms” of activity to others, generally of “higher” forms to “lower” forms. Modern science is full of these, which might also be called “genetic fallacies,” and has attempted — with obstinacy and sometimes even with ingenuity worthy of a better cause — to “reduce” aesthetic, moral, and religious activity to “biological” instincts, tendencies, needs, and interests, disregarding their essential and irreducible uniqueness.[xxiv]

A great number of scientific errors are then due to affective tendencies. However much they aspire to “objectivity,” scientists still remain — to a greater or lesser degree according to individuals and circumstances — subject to all the passions that disturb the judgment of people in general. There is certainly no need to insist on the enormous influence that feelings and passions have on opinions; it is a very ancient and common observation that one easily believes what one desires (quod volumus facile credimus[xxv]). So too it is useless to enumerate the tendencies, emotions, and passions that are causes of error; in various cases all of them can be. I will only mention that scientific activity for its own sake, by continually providing new facts and new conceptions, gives rise to particular affective reactions, due to the misoneism or neophilia[xxvi]which are present in greater or lesser degree in each person.

Finally, there are the professional errors of scientists, which depend on the particular mental attitude necessarily assumed by the devotees of each discipline, according to the character of that discipline. Renda drew attention to the psychologist’s errors in the first issue of Psiche:[xxvii] several times the “doctor’s error” (especially the psychiatrist’s) of exaggerating the importance of the pathological element in various manifestations of human life has been noted; the “errors” peculiar to the mathematician, the physicist, the jurist, the economist, and so on, have been spoken of, or could easily be spoken of.

What conclusions should we draw from this quick review of the main types of scientific errors? Perhaps a sense of skepticism toward science and distrust of scientists?

Of course not. We should not believe that the errors to which I have alluded are made only by scientists. Such errors (except for the “professional” ones that are neither the most frequent nor the most important) are committed all the time, and often in a far cruder form, by every educated and uncultured person. Indeed, the tendency to arbitrary generalizations and abuses of analogy, which are sometimes truly grotesque, is particularly strong in primitive mentalities. It is therefore generally a matter of mistakes made even by scientists because they are people; not of mistakes made only by scientists because they are scientists.

From this conclusion two, somewhat opposite, deductions can be drawn: 1. It is not right to be especially surprised and indignant because scientists make many and serious mistakes, much less to proclaim therefore the bankruptcy of science; 2. Scientists, on the other hand, should remind themselves more often how much they too are “people” and be more on guard against all causes of error. And to exercise this vigilance effectively, one must be well acquainted with the various types of possible errors.

A final remark. Not all errors are equally detrimental to the progress of science. On the contrary, as has been noted many times, some can succeed in being useful, either by giving rise to unforeseen discoveries or by provoking useful criticism and research.

In general, it can be said that among the most harmful errors should be counted simplistic theories, pseudo-explanations, and in general anything that tends to give a fallacious sense of security and conceal the complexity of issues. Also very harmful are many errors due to affective tendencies, and among these especially the misoneistic denials of new truths, which significantly slow down scientific progress.

More harmless, on the other hand, are certain arbitrary generalizations and certain overly bold and fantastic theoretical constructions. For while the exaggerations and errors they contain are easily uncovered and neutralized by the criticism of opponents, such constructions serve to highlight well previously neglected factors and can usefully stir the academic waters they threaten to swamp.

[i] Doc. #22324 is a copy of the original published essay, and Doc. #22323 is a hand-corrected typed manuscript of later date, clearly incorporating Assagioli proposed revisions to this essay – however many of the last pages are missing from this document. This translation comes from the psicoenergetica.com file which is clearly based upon the published essay. —Ed.

[ii] Editor’s interpolations are shown in [brackets] and elisions . . . are as shown in the original document. —Ed.

[iii] Francis Bacon (1561-1626) was an English philosopher and statesmen, author of numerous works, whose thought led to the advancement of natural philosophy and the scientific method. —Ed.

[iv] John Locke (1632-1704) was an English philosopher, physician and author, thought of as the father of empiricism. —Ed.

[v] Bacon, Novum organum sive indicia vera de interpretatione naturae. Lib. I, Aph. 39 ff. —Author’s Note.

[vi] Locke, An Essay Concerning Human Understanding, Book IV, chap. XX. Locke does not put errors due to language in this classification, but in his work he dealt at length with “abuses of words” (Book III, chap. X.) —Author’s Note.In describing “causes of error” In this chapter, Locke is describing what he calls “how men come to give assent contrary to probability,” which is why he lists the fourth reason as “wrong measures of probability.”  —Ed.

[vii] Histology is a branch of anatomy that deals with the minute structure of animal and plant tissues as discernible with the microscope. —Ed.

[viii] Julian Ochorowicz (1850-1917) was a Polish philosopher, psychologist, inventor, and scientist. —Ed.

[ix] “La tautologie expérimentale,” in Annales des Sciences psychiques 9 13, p. 29. —Author’s Note.

[x] Jean-Martin Charcot (1825-1893) was a French neurologist who did ground-breaking work in a variety of areas including hypnosis and “hysteria.” His ideas about hysteria were later refuted. —Ed.

[xi] The logical “fallacies of induction are all failures in reasoning about the messy world of cause and effect, contingent facts of the universe, and generalizations about kinds of things in the world. In each case, an argument is put forth using evidence incorrectly, or making bad predictions, or generalizing improperly.” (from an online course Thinking Well – A Logic and Critical Thinking Textbook, Chapter 9.5 “Fallacies of Induction by Andrew Lavin  —Ed.

[xii] Published in this journal [“La Psicobiologia come Scienza”]: Psiche II, 1913, p. 1). —Author’s Note.

[xiii] Prof. Wagner later published Das Zweckgesetz in der Natur: Grundlinien einer Meta-mechanik des Lebens [The law of purpose in nature: basic lines of a meta-mechanics of life] (1924) which put forth a biological philosophy that was not mechanistic. —Ed.

[xiv] Nicolas Malebranche (1638-1715) was a French priest and rationalist philosopher who sought to synthesize the thought of St. Augustine and Descartes. —Ed.

[xv] See La recherche de la vérité, Book II: De l’imagination. —Author’s Note.

[xvi] A good treatment of this topic, full of interesting quotations, can be found in G.A. Colozza’s book L’immaginazione nella scienza [Imagination in Science: Notes on Psychology and Pedagogy)] (Turin, 1899). —Author’s Note. Giovanni Antonio Colozza (1857-1943) was an Italian educator and author. —Ed.

[xvii] Jacques Loeb (1859-1924) was a German-born American physiologist and biologist, who was a teacher of John B. Watson, the “father of Behaviorism,” and whose work influenced B.F. Skinner. He was the author of The Mechanistic Conception of Life (1912), which has recently been republished in an annotated edition. —Ed.

[xviii] Loeb was a strict determinist who asserted that tropisms, or automatic, mechanical responses governed life, including behavior, fertilization and all human progress and mental activity. —Ed.

[xix] Martin Guelliot, in one of his very fine studies on the psychology of intelligence, brings together these and other errors under the general designation of “illusion d’expérience intégrale ou illusion de totalité” [“illusion of integral experience or illusion of totality], which he defines as “la conception comme un tout complet et intégral d’une partie arbitrairement découpée dans ce tout par les hasards [the conception as a complete and integral whole of a part arbitrarily carved out of this whole by chance]. Le Spectateur, III, 1911, p. 21. —Author’s Note.

[xx] “Historical materialism” is a theory of history, rooted in the philosophy of “dialectical materialism,” associated with Karl Marx and Friedrich Engels. —Ed.

[xxi] After all, the latter error is not always as gross as it might seem at first glance, since complicated actions and reactions between “causes” and “effects” often take place in life processes. —Author’s Note.

[xxii] Latin: “after this, therefore because of this.” —Ed.

[xxiii] Latin: “the most annoying idol of all!” —Ed.

[xxiv] The modern term for this fallacy is “reductionism.” —Ed.

[xxv] Latin: “We easily believe what we want.”

[xxvi] Misoneism: Hatred of what is new; neophilia: love of what is new. —Ed.

[xxvii] “Gli Errori dello Psicologi” by A. Renda in Psiche, Anno I. 1912 – Volume I, p.22. —Ed.