By Dr. Subhasis Chattopadhyay –
Artificial Intelligence and Research
Has the advent of AI made research easier? Or, does it pose more trouble for the researcher. Unscrupulous students and scholars now mostly use AI to hurriedly get over their assignments. There are AI engines solely dedicated to give the gist of a research-paper to the candidate and then the candidate often cites it in her paper, all the while pretending to have read the entire paper. This is fraught with dangers. Two examples will show how dangerous these AI summaries of research papers and books can be. Say one is working on a paper on the Prophet Jeremiah and somehow one has stumbled on Kathleen O’Connor’s book, Jeremiah: Pain and Promise, and then without reading the book, based on AI inputs one concludes that O’Connor’s is an exhaustive study of Jeremiah. But in reality O’Connor writes long about societal traumas which are universal in nature. As far as this book of hers is concerned she might be remembered as a trauma-studies expert rather than a Biblical scholar alone. This point can be and should be contested because these are this author’s own opinions. The point is that after reading O’Connor, one needs to sieve the material for the purpose of the research at hand. While she might aim at a Christian audience; her book appears to this author to expand into a universal study of the effects of trauma even in the twenty-first century.
Whether she should be mentioned in a research paper on Jeremiah is itself questionable. Perhaps she should be mentioned in the paper as an author whose insights make the sufferings of the Jewish people less informed by the Covenant and more diffuse applicable to all victims everywhere. This is precisely not the Judaic viewpoint. Another example, this time from the humanities will clarify the dangers of not reading an original source and yet referring to it in one’s research. Say somewhere one has heard that Graham Greene’s hard-boiled/detective fiction were like Agatha Christie’s novels and then one uses AI to generate a few comparisons; AI will give examples from both Greene and Christie. And then the researcher copy-pastes this and tries to pass off muster; this is the beginning of the researcher’s problems. Greene was influenced by G. K. Chesterton more than he was by Christie and some other later hard-boiled noir writers. Agatha Christie is too lowbrow for Greene. Greene drew his inspiration from Joseph Conrad when he wrote his thrillers. This can only be known the old, hard way — reading extant criticism on Greene and reading both Agatha Christie, Joseph Conrad, and the likes of Dashiel Hammett. This is demanding work. Bram Stoker’s novel, Dracula resists the Kantian Enlightenment project while accepting the categorical imperatives of Immanuel Kant is an insight which is this author’s intellectual property already published online; and no AI ever connected those dots.
Students and doctoral scholars often create bibliographies which they have never read. Sometimes these bibliographies are longer than the tutorial and the thesis, respectively. All these, just to show how much they have read. Yet, even the bibliography is generated by AI engines. Thus, within bibliographies errors creep in. The entire point of the research is lost. Instead of being an original work with a few unique, never before thought of ideas which integrate tradition with individual talent, in the sense that T. S. Eliot wrote about such integration, we have a total mess of innumerable so and so saying this and that which have little of value to add the research topic at hand. It is like watching a desert in a PowerPoint presentation in a super-specialised Bible studies’ seminar’s session on the Exodus event and Israel’s encounters with the Hittites. This author has had to bear such tortures as seeing the imagined face of Shakespeare in seminars on Shakespeare by Shakespeare-scholars. And the topic was ‘Hamlet as an existential anti-hero’. Notice in both cases, the presentations are AI powered hackneyed topics which the presenters have tried to pass of as original research. They had nothing new to talk about either the Israelites or about the Hittites. Similarly, generations have been stuck at existentialism and Hamlet. There is nothing original in this. AI even gets the proper quotes from any topics — with a catch. The quotes are more often than not de- contextualized. One last example will help to show the pitfalls of using AI indiscriminately for research: both Hindus and Christians when they research on Yoga, invariably come up with the standard etymological understanding that yoga is about a mystical union of the human person in the here and now with God. All sorts of AI engines throw up this definition. But the entire point is that yoga is a disunion which characterises the Samkhya yogi. Georg Feuerstein masterfully explains this understanding of disunion. Feurstein’s commentary on the Yoga Sutras have this note on yoga being a disunion. Since we are not reading anything deeply but skimming through information, we are missing these subtle points. There are distinct Indian forms of Catholicism and even Hinduism, as there exists a long distinctively Indian tradition of literary criticism of say, reading William Blake in India. AI simply does not know or care, for the correspondences between K. D. Sethna and Kathleen Raine.
The Research Problem/Topic
One researches because one has found a problem to address in the first place. If one does not find a problem worthy of re-searching then there is no point in authoring a dissertation. Every academic paper that is found online is not the product of research; neither are they always credible. They are often hastily written papers which are standard requirements for either getting a degree by research or are written helter-skelter for fulfilling promotion-requirements. Tenure-track faculty and even unfortunately, tenured faculty frequently churn out a lot of academic material which is a rehash of scholarly knick-knacks posturing as research. Often doctoral dissertations are just an extensive list of who said what and when. This cannot be a research problem. The point is to build upon the research of others. A good research proposal should not read like this:
The influence of Plato on St. Augustine of Hippo. This thesis aims to show how Plato shaped the thoughts of Augustine…
Rather, the research proposal could be something like this:
Since Plato’s influence on St. Augustine has been well documented, it is proposed to study how Plato’s thoughts affected Hannah Arendt since Arendt’s doctoral dissertation was on St. Augustine of Hippo. This research intends to study Arendt totalitarian state as a reversal of Augustine’s New Jerusalem and Plato’s conception of the ideal state…
Notice, in both cases, the language of the research proposal is lucid and avoids bombast. The proper research proposal is worthy of working on since till the time of authoring this essay, no such work exists which bridge the divide between political science and theology beginning from Plato. Very few are aware of Arendt’s dissertation. How can one study Western political science without knowing St. Augustine and his concept of the ‘polis’? AI knows all about Plato and St. Augustine separately, but they need to be connected by painstaking scholarship. AI can be used to for data mining. Beyond this, it is a waste of time to use AI in research. The larger question is what is gained by connecting Arendt’s theories of totalitarianism to Augustine’s understanding of the City of God. Augustine is a Church Father and Arendt is a Jew. How does one resolve the problem of their religious differences before proceeding to a study of Arendt? Also, how does Arendt challenge our views of Augustinian theodicy/theology? To what extent should scholars of Augustine recalibrate their readings in view of Arendt’s theories. Is it even necessary to read Augustine through a hermeneutical lens constructed out of Arendt’s theories? Is it possible that when we today speak of the Hospitable Other, we are rereading Augustine because it was Arendt who identified the Hospitable Other in Augustine in her doctoral work. Levinas scholars need to interrogate this. This author has come across discussions of Levinas without hearing anything of St. Augustine and Arendt. Research begins with what Sven Birkerts calls ‘deep reading’ in The Gutenberg Elegies.
Let us illustrate this with a few examples drawn from various domains. To be a Bible scholar, none needs to consult hundreds of Bible dictionaries in musty and remote libraries anymore. Access to powerful Bible software solves most of the scholar’s troubles. But what these search engines and AI cannot sort out is whether to use John Dominic Crossan’s insights regarding the historicity of Christ or, to decide to what extent to reject Crossan outright? Even if someone uses Crossan’s insights, then how does that enrich Christology? AI can give us summaries of Crossan’s insights, but AI cannot determine to what extent Crossan erred as a structuralist. The very statement that he erred in his academic work is obviously biased against Crossan will be something that AI can contest and yet this author knows through his Hindu Faith that Christ lived and was resurrected. This is the same Hindu Faith which tells this author that when Camille Bulcke researched and concluded that there was no historical evidence that certain core beliefs of Hinduism were true, Bulcke was entirely worng. As Crossan had thrown out the baby with the bathwater, so too Bulcke’s research led him to conclude that Sri Ram did not exist to begin with. Neither Crossan, nor Bulcke were insincere. Nor was Antonio Negri insincere in his decrying of the immanent in Western metaphysics. But all three; Bulcke, Crossan and Negri reached conclusions through their research which fly against the research of others in their fields and the research of these three was constrained by technological impairments and a simple predisposition towards their own group’s collective biases. So, while they did earn their research degrees and went on later to write important books, their research-results demand rethinking. The power of research is attested by the rereading of Sigmund Freud and Freud’s acolytes by Fr. Adrian Van Kaam. Van Kaam was not convinced that classical psychoanalysis was correct as shaped by Anna Freud, Melanie Klein and even, Sabina Spielrein. Every problem of the mind cannot be rooted in the flesh or, the Pauline sarx. Fr. Van Kaam did not agree with Jacques Lacan’s absurd theories which are a potpourri of everything including the set theory and linguistics leading nowhere. Thus, Fr. Van Kaam rejected most of Freud’s abracadabra and Van Kaam’s research led to the establishment of the discipline of formative spirituality. It is an entirely different matter that woke academics keep parroting the works of Jacques Lacan without often knowing Lacan’s involvement in the anti-psychiatry movement which led to the destruction of many lives. Had Lacanians not interfered, these patients might had been able to achieve euthymia and not progress to the clinical worsening of their symptoms. Though Fr. Van Kaam is not much known even within Roman Catholic circles, his was true re-search while the academic works of the likes of Crossan, Bulcke, Negri and Lacan, to name only a few will be remembered for their divisiveness and the harm that they did to humanity and how they continue to be used to brainwash gullible students. Research is not always proving something untrue but in building on what perhaps is unfinished work. Fr. Van Kaam finished the unfinished work of classical psychoanalysts without losing sight of self-evident truths. He did not succumb to logical positivism. There are, and were, millions of people, both Hindus and Christians, who have been called by name within both Hinduism and Catholicism to follow the Supreme Godhead in ways distinct from the ways in which the rest of humanity has been called by name by YHWH qua Brahman/Shakti. Fr. Van Kaam’s insights regarding this divine-human interface is now a new branch of studies: Formative Spirituality/Spiritual Psychology. His research led him away from established methods often thought to be etched in stone by both psychoanalysts and psychologists. The same was the case with Dr. Martin Seligman.
Seligman himself was trained in classical psychological methods and was well acquainted with Freud’s theories which he found to be empirically untenable. Seligman’s research led him to conclude that Freud’s construction of the psychic apparatus was too deterministic while lifespan psychology and motivational psychology both show us that we as individuals can and do change if we so want throughout our lives. We are not constrained by life-events which occurred when we were children. Thus, was born the new domain of positive psychology which is now taught globally with spinoffs even within theology — positive theology. Research takes us to new untravelled vistas. Otherwise, what is the point of research? It is untrue that doctoral research is just a training for greater things. Every bit of research adds up — no research can be the accretion of just what others including AI has said. There has to be constructive critiques leading to divergence if necessary. Divergence for the sake of divergence cannot ever be the norm of any research. One should not waste time by reinventing the wheel. Critiques are valuable only if they lead somewhere worth going. As far as religious studies and literature are concerned, one should never do doctoral work in subdomains in these subjects unless one is fluent in the languages in which the primary texts are written. One cannot be a researcher of Kashmiri Shaivism without knowing Sanskrit. Neither can one research Dostoevsky without knowing Russian.
Research problems cannot concern themselves with such naiveties like deciding whether Martin Heidegger was a greater philosopher than Edith Stein or whether Adi Shankaracharya was a more nuanced aesthete than Sri Avinavagupta. In the first case, Martin Heidegger was not interested in the Problem of Empathy as Stein was, and further Heidegger was more interested in the human person within the tradition of existential phenomenology. Stein, on the other hand, was concerned more with man’s relation with God in a way akin to the ethical positions of cloistered Catholic Christian mystics. What might constitute a research problem is to find out the exact location of the divergence of Stein from Heidegger’s thoughts. In the second case, Adi Shankaracharya was reacting to powerful Buddhist exegetes and, while he wrote beautiful verse, he focussed on proving that the world was real to the extent that Brahman qua God is real. Buddhist scholars had proven through their logic that the world is simply non-existent since the Madhyamaka (Madhyamika) way posits sunyata (‘voidness’) and dependent origination. Sri Avinavagupta, on the other hand, was a consummate aesthete who centuries before both St. John of the Cross and Hildegard of Bingen created an entire system of aesthetic reception rooted in scala spirituality within the Kashmiri Trika system. A more fruitful research problem may lead one to compare the conception of a powerful God by Adi Shankaracharya mediated through Sri Avinavagupta with that of the construction of God as weak and simultaneously powerful by the Church Fathers and later by the deconstructionist, John Caputo. Studying Advaita Vedanta alongside the Madhyamika way is now dated. Rather, one can look at how the Madhyamika affected Japanese forms of Buddhism and thereby influenced the Servant of God, Pedro Arrupe. These influences need working out for the very existence of our future which faces its greatest threat from AI. The way ahead is not superficial interreligious engagement but profound academic engagement with each other’s religions. This academic engagement will then trickle down to the faithful of different religions. Without this mode of interreligious scholarly research, we risk meaninglessly arguing with each other. This is with the caveat, that Shankara does not admit of an external God in the way that Christianity does. Here, one can refer to this author’s essay on the irrelevance of the category of atheism within Hinduism published at ESamskriti. It is unlikely that much will now be gained by studying Adi Shankaracharya’s conception of poetic metres which he often derived and revised from the Vedas. The point here is that the research problem has to be clearly identified and then addressed. And, it is not easy to identify research problems. A few more examples will prove this point.
When one studies Gerard Manley Hopkins’ poetry one might get stuck at his use of sprung rhythm. It is not sufficient to state that one needs to study Father Hopkins’ theology to understand his poetry. Nor is it sufficient to study The Spiritual Exercises of St. Ignatius of Loyola to answer the question why Hopkins chose sprung rhythm for his poetry. To begin with, are we speaking here of apophatic theology, or are we speaking here of a fallibilist process theology or, are we speaking here of Meister Eckhart’s influence on Father Hopkins or about none of these theologies but of the theology of the Jesuit scholars who shaped Father Hopkins’ spiritual life? This last line of thought takes for granted that just since Father Hopkins was a Jesuit, his intellectual formation was not say, Thomist in nature. Once one decides to read Hopkins as a theologian one has to search for the roots of his thoughts and that is another challenging but satisfying process. The more we go back in time, it becomes harder for us to find the exact sources that went into the formation of certain thought systems. Unless one has access to archival records or Hopkins’ letters, marginalia or class notes taken by him while he was at seminary, one cannot comment to what extent he was influenced by Patristics, Scholasticism, or the spirituality of his fellow Jesuits. Without evidence, everything is conjecture, and conjectures do not make for sound research. And the problem with Hopkins and sprung rhythm arises only if one is dissatisfied with the answer that this poetic metre is the closest to natural sounds/the spoken idiom. Hopkins scholars need to think whether it is even worthwhile to research sprung rhythm anymore through theological lens. This proves that formulating research problems may be easier than justifying why such problems need to be resolved at all.
Now, this brings us to the need for accessing credible sources while conducting our research.
Just because an article is found online, even if it is on JSTOR, it does not establish the article’s or author’s credibility. Who is the author of this article; can the author be relied upon to be honest in her understanding of the topic dealt with? This is crucial. There is no point in selecting Jane Doe as a reliable source of information since Jane Doe is a prolific writer.
On the other hand, when one is an English Major in India, one can safely begin by reading Bhabatosh Chatterjee on John Keats, W.B. Yeats, and Bankim Chandra Chattopadhyay. It is no mean feat that Tarak Nath Sen did not travel out of the country and first taught at a rural college and yet wrote his paradigm-shifting work on Shakespeare with scarce resources. How does this author know that Chatterjee and Sen are important for researching Keats, Yeats, and Bankim Chandra Chattopadhyay who composed our National Song and wrote Anandamath? One might blindly accept that Ferdinand de Saussure’s theories about the arbitrary relation between the signified and the signifier is etched in stone, but for all we know, this might have been a strategy on the part of Saussure to negate Hindu, Buddhist and even Christian understanding of the power of the Logos or, Word. In fact, the relationship between the signifier and the signified is never arbitrary. Saussure was a Sanskrit scholar, and he knew this truth. So, we must be aware of standard works when we begin a research project. Kwame Appiah did not find Cosmopolitanism only in the Cynics. He elided any reference to theories of proto cosmopolitanism found in ancient Jain texts. On the other hand, Giorgio Agamben sincerely tried to understand Shaivism, though he was wrong in his analysis since he is not conversant with Sanskrit. These elisions and appropriations should be well understood by scholars; otherwise the research will always be intellectual sus. Within the context of Biblical studies, if one were to re-search the Johannine corpus, one is struck by the simplicity of Father Raymond Brown’s understanding of the non-synoptic Gospel of John. How can one be a Johannine scholar without knowing anything about Raymond Brown? One builds on the foundations of Fr. Brown’s profound insights on John’s Gospel. Similarly, one cannot know anything of Homer today, without knowing about the Milman-Parry thesis. To research, for instance, the Upanishads, Swami Gambhirananda’s glossing of Adi Shankaracharya’s exegesis of the major Upanishads remains the gold standard in Hinduism. Both Sarvepalli Radhakrishnan on the other hand, like Surendranath Dasgupta have slightly dated. They should be read only for their historical values as pioneers within the presentation of Hindu philosophy to an English-speaking audience. But to really understand say, Samkhya, one has to turn to Swami Hariharananda Aranya. It is impossible for Aranya to have read Plato; yet both Plato and Aranya opine that poetry being unreal should be avoided. Both miss that literature makes truth-claims and does not lend itself to simple categories like right or wrong. This is because while the entirety of Western (Christian) philosophy is Aristotelian; admitting of ‘the excluded middle’, when Plato formulated his theories, Aristotle’s works were not there. And Indic philosophies have never excluded the middle. The connection between Plato on literature and Samkhya on literature is this author’s observation. It can be elaborated by interested scholars but had this author depended on AI and did not deeply read Plato and Aranya on Samkhya, he would not have been able to make this crucial connection. This is the purpose of research — it is not about rehashing what others have said or arrogantly rejecting others; it is all about building on the research of others. Research that pretends to be sui generis is simply not research.
Citations and why they are important:
The obvious point is that without giving credit to an author, one gets blacklisted as a plagiarist. The more critical issue is that a citation helps scholars interested in an idea, to research further about that idea. Often, citations are hurriedly done using various online citation machines and following the rule of the law more than the spirit of the law. The idea is to provide as much information as one can to help future scholars. No matter which citation method one follows, the citations should be detailed. For example, when we quote Sigmund Freud, we should always mention the edition and if we use websites to access his works, then too we should spend time finding the name of the translator. Without the name of the translator, the entire reference to Freud becomes useless. A translator may agree with Freud or disagree with him, and that skews the translation. Thus, a researcher might want to avoid a certain edition of Freud in preference to another edition. It is better, if one is a serious scholar, to annotate the bibliography so that generations hence, yet unborn scholars know what was in that web page, or online forum which was consideration-worthy. This problem was not there when there was no internet. Web pages are ephemeral. So, the need for copious annotations arise. Neophytes are bogged down by citation methods and fill bibliographies with articles they never read; detailed annotations also prove the integrity of the scholar. Nothing should be done in a hurry; even though there are absurd deadlines to be met. One should think more, read continually, and then write. Otherwise, that machine generated bibliography is all hoodwinking. A good bibliography might have a few formatting errors. But they never have a dearth of information. And there are smart-Alecs who fill their bibliographies with books they have never read without considering the fact that most of us no longer refer to hardcopies. We use academic databases. So why lie? Further, if one is referring to books all the time, perhaps one is outdated and is going to invent wheel again in her research. Dictionaries online are being continually updated. Why not use them rather than pretending to use an old dusty volume which is already online? One suspects the intentions of the scholar are malafide and is to trick others. This is why so much of the research output in India is puerile, unoriginal, and not worth anyone’s time. .
Subhasis Chattopadhyay Ph.D. is an ex-Sahitya Akademi Judge. His Ph.D. was on theodicy and Patristics in Cormac McCarthy and Stephen King. He has formal qualifications in Biblical Theology and in Formative Spirituality. He has also formally studied the Hindu canon and has separate qualifications in the behavioural sciences. His reviews in Prabuddha Bharata from 2010 to 2011 have often featured in the websites of Ivy League Presses. His works in various fields are much cited. He writes here and at ESamskriti and the Herald. Dr. Chattopadhyay delivered the de Nobili Endowment Lecture in 2022.