If you want to take Tolkien research seriously, you have to set some standards for yourself and abide by them. There are no generally accepted standards of research, particularly not for the online Tolkien fan community. This guide only provides suggestions that should be easy to understand and work with.
Tolkien research can be divided into several categories: traditional literary criticism (explaining what the author is trying to convey to the reader, or what the author wants the reader to feel); deconstruction or reverse-engineering Tolkien’s stories (identifying possible sources and roots for the various characters, plots, modes, etc.); linguistic research into Tolkien’s invented languages (which can be further sub-divided into studying potential source languages and models AND studying the structure and use of the invented languages themselves); historical research into Tolkien’s invented histories (the movements of peoples, the formations of tribes and nations, the tabulation of events and trends, etc.); research into the invented cultures (identifying specific customs peculiar to each group, identifying shared customs and ideas, etc.); and the study of Tolkien studies.
One can certainly strive to become a master of all these trades, but you’ll succeed most easily by choosing one area and focusing on that until you are ready to move on. You don’t have to commit everything to memory. You don’t have to have all the books. It certainly helps to have both an encyclopedic memory and a large library.
Which area you begin with should really be a matter of personal preference. In many cases, the source materials are useful in several areas. For example, many of the articles published in the Vinyar Tengwar journals help both the linguists and the historians, and occasionally the anthropologists can gleen something about the cultural values of the invented peoples from the invented language rules. (“The Shibboleth of Feanor” provides a wealth of information regarding Noldorin history and culture in Aman, for example, though its primary function is to explain an arbitrarily practiced phonetic shift in Noldorin Quenya).
You also have to sort out your source materials, establish some rules about what you will use and what you will not use, and determine how flexible you’ll be. This site, for exmaple, will never recommend the use of popular online resources such as The Encyclopedia of Arda, various FAQs, or the Wikipedia. They are, at best, poor sources of information and at worst misdirectional sources of information. The tendency to insert personal visions and political stances into online resources is strong among Tolkien fans.
Your research should be based first and foremost on the Tolkien books. But even so simple a suggestion immediately opens the door to questions of canonicity. Which Tolkien books are authoritative? Technically, they are all authoritative — but they cover different aspects of Tolkien’s career.
Even the novels and compilations — the “core” books, as it were — are peppered with multiple editions, secondary editorial decisions, typographical errors, and other incompatibilities. There is no true canon with respect to Tolkien’s fiction. Hence, there can be no single, universally correct or acceptable canon with respect to the study of his works.
In the introduction to Understanding Middle-earth, I explained how I arranged the source books I used for the essays in my books in degrees of authoritativeness. While my system is no more popular than anyone else’s (so far as I know), it is at least coherent.
For one thing, I limit my study of “Middle-earth” to the world established by the works published by J.R.R. Tolkien within his own lifetime. That precludes the use of information from sources like The Book of Lost Tales, The Fall of Numenor, and the first edition of The Hobbit. However, it complicates the use of materials that Tolkien himself prepared in anticipation of the publication of The Silmarillion, as well as other post-LoTR materials which eventually Christopher Tolkien eventually published in Unfinished Tales of Numenor and Middle-earth.
Many people, including some whom I feel should know better, resort to canons of convenience. That is, in order to deal with the complexities of the post-LoTR texts, they incorporate pre-LoTR texts into their source materials and treat everything as if it were coeval, designed and intended to convey information about the same fictional worlds.
The distortion of Tolkien’s creativity for sake of one’s personal convenience is inexcusable and certainly adds no value whatsoever to the study of any of Tolkien’s mythological ideas. Middle-earth, as described in The Lord of the Rings, did not exist in Tolkien’s published imagination prior to 1954 — and technically was not complete until 1955. The pre-1940s texts that many people use as resources do not anticipate, incorporate, or in any way cooperate with the LoTR-era and post-LoTR inventions of myth, language, and culture that Tolkien devised.
Nonetheless, as the linguistic community have shown, it is possible to extrapolate from the basis of earlier writings potential paths of development. There is no fundamental error in seeking such extrapolation. The errors of judgement occur when any one path of extrapolation intermixes multiple sources, or presents itself as more likely than any other path of extrapolation without demonstrating that multiple paths can lead to the same or very similar conclusions.
The fundamental rules Tolkien followed in devising his languages did not change much. But the axioms of his mythologies were asserted only to create specific perspectives. Hence, The Fall of Numenor provides no insight into the history of Numenor as it is expressed in the Middle-earth mythology (the mythology which truly began with the writing of The Lord of the Rings).
However, The Fall of Numenor provides a glimpse into the early development of Tolkien’s conceptualization of an “Atlantis-story”. That conceptualization itself is external to any mythology, and follows its own evolutionary path to the nearly full expression provided in works such as “Akallabeth” (published in The Silmarillion), the “Numenor” essay (published in Unfinished Tales of Numenor and Middle-earth), and stories like “Aldarion and Erendis” and “The Disaster of the Gladden Fields”.
These works clearly form part of the continually evolving Middle-earth mythology, but they have no part or place in the previous Numenor or Beleriand mythologies. The echoes of the older works resonate throughout the more fully mature ideas and themes of the finished or near-finished works, but those works are set in the world of The Lord of the Rings, and they belong solely to that world.
Hence, the study of the linguistics of Middle-earth must take into consideration the changes Tolkien implemented in the course of devising Middle-earth. But the study of the history and cultures of Middle-earth must restrict itself only to the works which form the mythology itself. And yet, the study of the sources for the stories and the mythological frameworks may freely move among Tolkien’s various mythologies, drawing comparisons and showing the evolution or stability of various borrowings from literary and historical antecedents.
If your goal is, say, to study the Middle-earth mythology itself — to acquire a deeper understanding of that world as Tolkien may have viewed it and sought to describe it — you cannot look at the pre-LoTR texts. They provide no foundation, nor any insight, nor any degree of authoritative reference for indexing the basic data that comprise the fictional world. There are no Gnomes in Middle-earth. Tevildo Prince of Cats did not serve Morgoth. And Morgoth did not send a thousand wingless Balrogs riding across the plains of Ard-galen in the wake of Glaurung.
That is really where the study of Tolkien’s Middle-earth breaks down into the study of a myriad of possible Middle-earths, each of them unique, many of them similar to both Tolkien’s world and each other, and none of them authoritative or reliably usable in understanding Middle-earth.
The careful Tolkien researcher refrains from saying “This could have been possible” with such force that it becomes compelling to infer that “if it is possible, it is probable”, and thus to conclude that “if it was possible and therefore probable, then it is most likely”. The majority of casual commentaries follow this path of self-vindication, and they are no more valid in terms of understanding or explaining Tolkien’s work than blubbering to ones’ self in the bathtub.
Virtually anything Tolkien did not write is in some degree a possibly correct or logical extrapolation. But extrapolation itself cannot answer the question of what Tolkien intended or where he would have gone. It can only define (hopefully) reasonable limits. One sees the breadth of the scope of Tolkien’s vision more clearly by ascertaining the probable boundaries of that vision.
But Tolkien was in many ways unpredictable. He never wrote down his rules of sub-creation. We glean a few possible ideas here and there. His philological work clearly helps linguists understand how the languages should evolve and progress. But human history and human culture are the practical outreaches of a multitude of unique individual expressions and preferences. Your research can never divine the exact progression Tolkien would have followed in any provable sense.
In the end, you have to settle for, “We don’t know what he would have done”, and hope that what you have proposed at least makes sense to your audience. The research in itself can be interesting if not entirely useful. But Middle-earth is hardly a stagnant backwater of irrational imagination. It flourishes and grows today. Let that growth guide your research.
You don’t need to seek vindication. Just be sure you are consistent in applying your rules. And if you’re not sure of what your rules should be, write them down. You may not get them right the first time, but the more often you try, the clearer your path will become.