What Does It Mean to Be Scientific?

Some identify science with natural sciences or quantita­tive research. In other words, they believe research is only scientific if it contains formulas and diagrams. From this perspective, research on Aristotle’s ethics would not be sci­entific, nor would a thesis on class consciousness and the peasant revolts during the Protestant Reformation. Clearly this is not the meaning that academia assigns to the term “scientific.”

Let us try to understand by what reasoning we can call a work scientific. We can still take as a model the natural sci­ences as they have been defined since the beginning of the modern period. In this sense, research is scientific when it fulfills the following conditions:

1. The research deals with a specific object, defined so that others can identify it. The term “object” need not neces­sarily have a physical meaning. Even the square root of a number is an object, though it cannot actually be seen or touched. Social class is also an object of research, despite the objection that we can only know individuals or sta­tistical means and not actual classes. In this sense, the class of all integers above 3,725 also lacks physical real­ity, though a mathematician could study it. Defining the object therefore means defining the conditions by which we can talk about it, based on rules that we establish, or that others have established before us. If we establish the conditions that allow anyone to discern an integer above 3,725 when he encounters it, we have established our object’s rules of identification.

Obviously, problems arise if we must speak, for exam­ple, of a fictional being such as the centaur, commonly understood to be nonexistent. At this point we have three alternatives. First, we can decide to talk about centaurs as they are presented in classical mythology. Here our object becomes publicly recognizable and identifiable, because we are dealing with the texts (verbal or visual) in which these mythical creatures appear. We will then have to determine the characteristics that an object being described in clas­sical mythology must possess for it to be recognized as a centaur. Second, we can conduct a hypothetical investi­gation to determine which characteristics a creature liv­ing in a possible world (that is, not the real world) should possess in order to be a centaur. Then we would have to define the conditions of existence of this possible world, taking care to inform our readers that all of our discussion is developed within this hypothesis. If we remain rigor­ously faithful to the initial assumption, we have defined an object appropriate for scientific investigation. Third, we can produce sufficient evidence to prove that centaurs are in fact real. In this case, to build a realistic object of discussion, we should present evidence (skeletons, bone remains, tracks petrified in lava, infrared photographs from Greek woodlands, and whatever else might support our case) so that others might agree that, regardless of the correctness of our hypothesis, there is something we can talk about. Obviously this example is paradoxical, and I can’t believe that anyone would want to write a thesis on centaurs, especially by way of the third alternative. Instead, my purpose is to show how it is always possible, given certain conditions, to constitute a publicly recogniz­able object of research. And if it is possible with centaurs, it will surely be possible with notions such as moral behav­ior, desires, values, or the concept of historical progress.

2. The research says things that have not yet been said about this object, or it revises the things that have already been said from a different perspective. A mathematically correct the­sis that proved the Pythagorean theorem with traditional methods would not be a scientific work, because it would not add anything to our knowledge. At best, it would provide clear instruction on how to solve the theorem, much as a manual provides instruction on how to build a doghouse using wood, nails, a plane, a saw, and a ham­mer. As we have already said in section 1.1, a literature review can also be scientifically useful because the author has collected and organically linked together the opinions expressed by others on a particular topic. Similarly, an instruction manual on how to build a doghouse is not a scientific work, but a work that discusses and compares all known doghouse-building methods can make a mod­est claim of scientific value. However, bear in mind that a literature review has scientific value only if something similar does not already exist in a given field. If someone has already written a work comparing the systems used to build a doghouse, writing a similar manual is at best a waste of time, at worst plagiarism (see section 5.3.2).

3. The research is useful to others. An article that presents a new finding on the behavior of the elementary particles of physics is useful. An article that presents a transcription of an unpublished letter by the Italian romantic poet Gia­como Leopardi, and that recounts the circumstances of its discovery, is useful. A work is scientific if, in addition to fulfilling the two conditions above, it advances the knowl­edge of the community, and if all future works on the topic will have to take it into consideration, at least in theory.

Naturally the scientific relevance is commensurate with the contribution’s significance. Scholars must take certain contributions into account in order to say any­thing relevant on a particular topic, while they can leave others behind without serious consequences. Recently, a number of letters from James Joyce to his wife have been published, specifically letters that deal with explicit sexual matters. People studying the origin of Molly Bloom’s char­acter in Joyce’s Ulysses may find it useful to know that, in his private life, Joyce attributed to his wife a sexuality as vivacious and developed as Molly’s. Therefore, the publi­cation of these letters is a useful scientific contribution. On the other hand, some superb interpretations of Ulysses present a keen analysis of Molly’s character without this data. Therefore this contribution is not indispensable. We can find an example of a more important scientific contribution in the publication of Stephen Hero, the first version of Joyce’s novel A Portrait of the Artist as a Young Man. Stephen Hero is generally considered fundamental for understanding the development of the Irish writer, and is therefore a fundamental scientific contribution.

Here we should address the so-called “laundry lists” often associated with extremely meticulous German phi­lologists. These might include an author’s shopping list, to-do list, and other incidental texts that are generally of low value. Occasionally these kinds of data are useful because they shed the light of humanity on a reclusive author, or they reveal that during a certain period he lived in extreme poverty. Other times these texts do not add anything to what we already know. They are small biographical curiosities with no scientific value, even if there are people who build reputations as indefatigable researchers by bringing these trifles to light. We should not discourage those who enjoy pursuing this type of research, but we also should understand that they are not advancing human knowledge. From a pedagogical perspective, if not from a scientific one, it would be more fruitful for them to write an entertaining popular biogra­phy that recounted the author’s life and works.

4. The research provides the elements required to verify or disprove the hypotheses it presents, and therefore it provides the foundation for future research. This is a fundamental requirement. For example, to prove that centaurs live in Peloponnesus I must do the following with precision:

(a) produce proof (as we have already said, at least a tail bone); (b) recount exactly how I discovered and exhumed the archaeological find; (c) instruct readers on how more evidence can be unearthed; and (d) if possible, give examples of the precise type of bone (or other archaeo­logical find) that would disprove my hypothesis, were it to be discovered in the future. If I accomplish these four goals, I have not only provided the evidence to support my hypothesis, but I have facilitated the continuation of research that may confirm or challenge it.

The same is true for any topic. Suppose I am writing a thesis on an Italian extraparliamentary movement that took place in 1969, and that is generally believed to have been politically homogeneous. In my thesis, I wish to prove that there were in fact two factions, one Leninist and the other Trotskyist. For my thesis to be successful, I must produce documents (flyers, audio recordings of meetings, articles, etc.) that verify my hypothesis; recount the circumstances of the acquisition of this material to provide a foundation for further research; and present the criteria by which I attribute the supporting documents to the members of the 1969 movement. For example, if the group was dissolved in 1970, I must weigh the rele­vance of material produced by members while the group was active against that produced by former members of the group after its dissolution, considering that they may have cultivated their ideas while the group was still active. I must also define the criteria for group membership, such as actual registration, participation in meetings, and pre­sumptions of the police. In doing this, I provide the foun­dation for further investigation, even if it may eventually invalidate my own conclusions. For example, let us sup­pose that I consider a person a member of the group based on evidence from the police, but future research exposes evidence that other members never considered the person in question as a member, and therefore he should not be judged as such. In this way, I have presented not only a hypothesis and supporting evidence, but also methods for its verification or falsification.

The various examples that we have discussed demonstrate that a student can apply the requirements for scientific valid­ity to any topic. They also illustrate the artificial opposition between a “scientific” and a “political” thesis. In fact a political thesis can observe all the rules necessary for scientific validity. For example, I could write a thesis that is both scientific and political, and that would analyze my experience as an activist establishing an independent radio station in a working-class community. The thesis will be scientific to the extent to which it documents my experience in a public and verifiable manner, and allows future researchers to reproduce the expe­rience either to obtain the same results or to discover that my results were accidental and not linked to my intervention, but to other factors I failed to consider. The beauty of a sci­entific approach is that it does not waste the time of future researchers. If a future researcher is working in the wake of my scientific hypothesis and discovers that it is incorrect, my initial hypothesis has still proven useful. In this exam­ple, if my thesis inspires a future researcher to also become an activist in a working-class community, my work has had a positive result, even if my original assumptions were naive.

In these terms, there is clearly no opposition between a scientific and a political thesis, and as we have seen, one can write a “scientific” thesis without using logarithms and test tubes. On one hand, every scientific work has a positive political value in that it contributes to the development of knowledge (every action that aims at stopping the process of knowledge has a negative political value); but on the other hand, every political enterprise with a chance of success must be grounded in the scientific diligence I have described.

Source: Eco Umberto, Farina Caterina Mongiat, Farina Geoff (2015), How to write a thesis, The MIT Press.