Every conversation, argument of import, runs into the same confusions of meaning. We all know how the meanings of words can shift with context. We all know how one man's meaning of a given word can differ substantially from another's. I'm not contributing anything new by commenting on this.
I have some elaborations on this idea that I believe can be helpful. These are inspired by, or taken directly from Gilles Deleuze's work in a book called "What is Philosophy."
When I was growing up in my liberal, Enlightenment-revering, and Rationalism-revering household, I came to see Science as the gold-standard, and standard-bearer of truth. If pressed, I would have admitted to hearing truth of sort in a beautiful song, or a novel, but I would have rationalized it away as mere reflections, accidental revelations of truth by gracious accident. The only source of rigorous truth was science, as far as I was concerned. There was no limits to the prospective scope of science. Consciousness, the soul, the all-in-all, everything divine was accessible at least in principle to science. Ultimately, everything in the universe could, at least in principle be explained by science. The nice thing about science is its bulletproof approach to definitions (upto Kurt Godel), hard reference. In science, as in mathematics, words do not change meaning by context. The difference between the hard and 'soft' (non) sciences are seen in this. The definition of "accelleration" is static, while the definitions of "fascism" are myriad.
At some point I was reading Noam Chomsky's natural philosophical discourses in the Anshen lecture series publication "Language and Thought". One of his main points was that in the history of science, successes in reductionism (the manic objective of neurotic atheistic materialist reductionists) were quite rare. On the contrary, what was successful was unification. Unification instead of reductionism, but of what? Of systems of abstraction. Chemistry has a certain system of abstraction. Biology is above that, and molecular physics is below it. Each of these systems is successful at explaining and predicting phenomena/observations at their respective levels of abstraction. Chemistry hasn't been reduced to or replaced by molecular physics so much as unified with it. Biology hasn't been eliminated in favor of a chemistry-oriented ontology of living processes, but it has been unified with it. I don't want to belabor this point, I highly recommend this lecture publication and his other, a proper book, called "Language and Problems of Knowledge." Arguably Chomsky's theoretical contributions to linguistics don't amount to much, and his political commentaries are irrelevant here, but I do consider him an absolute treasure of wisdom in terms of modern natural philosophy, based on those two readings alone.
Interestingly, on the subject of psychology, Chomsky observed that science doesn't even have bragging rights on the truth in this domain. He argued that one could gain vastly more insight of human psychology through the reading of this or that classic novel, or this or that history, than by the most careful reading of the latest cognitive psychology theoretical treatise.
So by this time I recognized intrinsic limitations to the efficacy of science, although it still had a visible advantage in the matter of rigor and stability of meaning and definition.
A decade later I happened upon the aforementioned book by Deleuze, where he (and co-author Felix Guattari) offer up a kind of holy trinity of intellectual activity: Science, Art, and Philosophy. None superior over the others. Each of them methodologies by which man may do battle with chaos, extracting a place of reprieve, of apparent knowledge, shelter from the storm of abject confusion. Science is built on a "plane of reference", philosophy is built on a "plane of consistency/immanence", and art is built on a "plane of sense". These planes differ in characteristic dynamics.
These planes of reference that Deleuze refers to are corellated with what Chomsky would call a level of abstraction. So (classical) physics has a plane of reference, on which it would put objects/definitions/functors like "force", "gravity", "mass", and suchlike. Chemistry would be another plane of reference, and biology still another, although to the extent that two planes were unified or reduced one to another they would be then of one plane of reference. The point of calling it a plane of reference is that the reference is fixed, perfect. In science the objects inhabiting the plane have fixed reference, static meaning. Those objects are called functions, or more generally functors. Briefly, the plane(s) of immanence of philosophy are inhabited by concepts. Concepts are the objects inhabiting the plane(s) of immanence of philosophy as functors are the objects inhabiting the plane(s) of reference of science, and percepts iirc are the objects inhabiting the plane(s) of sense of art. Deleuze goes into a great elaboration of these ideas, differences in dynamics and properties of these tripartite opposed systems of intellectual construction. I just wanted to outline the basic idea. Functions on planes of reference, concepts on a planes of immanence, and percepts on planes of sense: science, philosophy, and art, each ways of constructing facsimiles of real truth, none fully apprehending that truth, each construction only provisional and tentative, awaiting its destruction and surpassing by more advanced constructions, ad infinitum. Man is limited of course...
Deleuze says something elegant: "Concepts are all signed." Signed by the thinker that created them. Chomsky alludes similarly to the personal characteristic of scientific concepts ("functions" in Deleuze's system, or formal ontology to be slightly more precise). This is the main point I was trying to reach in this piece, in order to arrive at a solid foundation of discourse. Certainly there is objective truth, and truth matters. Words on the other hand, intrinsically are mere tokens, and while in some cases we may arrive at consensus on their meaning (in a physics setting this is typical), in matters of lived experience there are domains where the meanings of words is not fixed at all, though they could be, there may be no authoritative definition of a word like "fascism" or "capitalism" or "narcicist" in a given domain of discourse. In a discourse, you need to define your terms yourself, you create (perhaps having first appropriated them from elsewhere) your terms, and you sign them. Whether you are an artist operating in an artistic discourse, on a plane of sense, and you sign your work, it bears your signature one way or another. Perhaps you proceed rigorously, carefully, painstakingly so as to have perfect reference, you define your plane of reference and sign it, your functions are yours (perhaps appropriated, but certainly modified) and you sign the whole thing. Maybe you are playing fast and loose with your intellectual processes, you are in the realm of philosophy, creating concepts, which unlike scientific functions are like quicksilver: "every concept is a cipher, it is multiple components, and each of these components is in turn a cipher, multiple"; these concepts are signed intrinsically by the one who thought them.
So the foundation of discourse I'm proferring is a maxim: Any proposition you put forth, don't assert normativity in the meanings of the components/words of those propositions: "Everyone knows such and such is communism". Instead, claim owneship (while extending credit to prior owners) of all the components of your propositions. You want to talk about capitalism? Fine, but first define it. And better yet, define the methodology of the thought you are engaging in. Is this science? Well, your definition better be airtight, on a self-enclosed, coherent system of reference in which your the components of your definition are themselves defined on that same plane of reference, and so on ad-infinitum. If you proceed philosophically or artistically, the rules may be different, but always claim ownership of the ideas you are constructing with. In this way, we won't "talk past each other", and we won't fall into the trap of condemning anothers constructions, their tools of thought.
So, that's the end of that piece. However, there is a related, sort of corollary to this, I wrote about some years ago, which I would like to take the opportunity here to have another go at. To wit, why are the "soft sciences" 'soft', and could they be made 'hard', i.e. rigorous? If so, how?
Those things above we called planes of reference are what makes a science. There are no soft sciences. Science is hard, i.e., it's reference is stable, meaning is fixed and unambiguous. Economics, outside of rigorous micro-specialities, is not a science. Sociology is not, in general, a science, though this or that thinker may make substantial progress in establishing for themselves and perhaps a few cohorts a small plane of reference on which to stand; in general, there is no shared plane of reference for sociology, so, at best, it's a science in utero. Before proceeding, let us digress a bit into a natural question that should arise at this point. What is history? What is anthropology? Are these sciences, arts, philosophies? Here and there, they contain elements of all of them. Some historians do their utmost to establish a plane of reference, to fix rigorous definitions, but their task defeats them by scale. Nevertheless, there is in principle a way of doing history as a science, states of affairs on a plane of reference. It may be more fruitful to proceed faster and looser, some histories proceed more philosophically, and in historical novels clearly there is transition even to artistic constructions and methods of approaching truth. Anthropology, in the academy may likewise proceed as much as possible adhering to scientific rigor, yet in practice the most fruitful anthropological fieldwork can be a dive into the unknown, experiential art, sometimes even pure fiction which nevertheless captures the heart of a thing more elquently and intimately than could any scientific detachment (Castaneda?). All of this is to say, although we love science and want to extend real science into these domains, we do not propose the annhilation of applications of these para-scientific methodologies to these domains; rather, we seek to put the scientific approaches on better footing, that they may proceed more fruitfully.
So, how to do this?
Science is a built with planes of reference, self-contained discursive systems of stable reference, stable definitions. This could also be called a formal ontology. Science fails in a discourse when the participants don't share a common frame of reference, plane of reference, formal ontology, (all synonyms here).
What computer software enforces is precisely a stable plane of reference, a stable and fixed formal ontology. The more well developed, expanded, intricate, complex, and sophisticated is a plane of reference / formal ontology, the greater the science.
While in general sociology might be unscientific, without shared formal ontology, I can give you an example of a perfectly scientific sociological framework: SimCity. Of course, it may be inane, it may be weak, it may not explain much, and it may be false, and forget about the graphics, we are focusing on the state model, it is perfectly well-defined. SimCity has everything needed to make a scientific system in terms of the stability of definition of the terms of the model of the city and society. The only thing left to do is make it better. Therefore, my proposition was (I first thought of this in 2010, and wrote about it on some Anthropology webblog), and is, sociologists should be teaming up with software developers to make game/models of their conceptions of reality as pertains to their formal ontological systems / paradigms of sociological reality. The can model a human, and define human interactions, see how they form emergent properties like communities, do experiments in the simulation world, adjust, refine, expand, and so on. Sociologists from all over the world could cooperate on massive systems like this, and competing schools could fashion competing systems. In this way, perfectly scientific sociological systems would be produced, where the meaning of primitive terms would be absolutely fixed, as would the meaning of emergent terms like "capitalism", "communinism", etc. So, we would have succeeding in making sociology a hard science, i.e. making it a science, making it scientific.
That's basically it. There a couple of minor points I want to make, as a postscript, some sort of lemmas:
People have a misperception of what science is, there is something of a deception going on about this. They think it's related to slogans like "empricism", "data", "experiments". Actually, art has experiments, but that doesn't make it science. Life can be full of experiments. Data can be meaningless if it isn't intepreted intelligently. Many "studies" are done with fatuous thinking. In academia many charlatans get tenure doing "scientific studies" which perform various "experiments" and then "analyze the data", and come to "conclusions." There is a lot of bullshit like this. Certainly experimentation is a critically important tool to arrive at truth in scientific matters, but as we have seen, experimentation is a tool in many aspects of life, including those having little or nothing to do with science.
The second point: An alert reader may have noticed that one could make a perfectly stable plane of reference, a perfectly static formal ontology, that has nothing to do with reality. Am I claiming that this fantasy world (maybe Legends of Zelda) is a science? Good question. So, actually yes, according to our definition, strictly speaking, it is a science, but it doesn't pertain to our world. It pertains to the possible-world of The Legend of Zelda. It's one thing to make a formal ontology, it's another thing on top of that to make a formal ontology that models well our world. This is the challenge of science. And of course this is where experimentation comes in.
With respect to the formal-ontologies required as a necessary defining condition for a science: they need to be complete, and comprehensive; they need to cover completely the entirety of the propositions asserted, all of the components of those propositions, and decomposing those components, must go down to the bottom (it's turtles all the way down). I.e. Every component of a proposition must be well defined in the ontology, and to the extent that those can be decomposed, all of it must be defined under the same formal ontology. So, the whole system is self contained, covers the entirety of the discourse intended, i.e. is complete and comphrehensive.
Notwithstanding Kurt Godel, who apparently proved that this was impossible in an absolute and airtight way, for a complete science. I won't pretend to understand his work, but am happy to say that clearly math, physics, chemistry, are good enough. What is not good enough are tiny little incomplete patchwork formal-ontologies, where an academic (let's say a macro-economist) makes a few grandiose definitions -- and definitions are a special type of proposition -- while neglecting the responsibility to decompose the components of that definition, and define as well, those terms, rigorously, the whole thing in a consistent and complete way.
It really is an impossible task for one scholar, but worldwide teams of like-minded theoreticians, working with a dedicated software engineering team on a game/sim, it is a feasible project. The formal ontology would be expressed in the game/sim, in the state model.