Hello everyone,

I decided to make this chapter FREE since it pertains to literary criticism, and I know there are a lot of literary folks on Substack.

Thanks for reading!

—Neal and Tina

“Unchecked generosity, because it wants to maximize the sense, intelligence, and truth of authors, will want to attribute unity to the largest possible whole.”

14

Generosity Beyond the Sentence

MEANING DEPENDS ON CONTEXT, and context is a series of ever-more-encompassing wholes. The word-sentence relationship is but one part of the series. Below it are the mere sounds—prefixes and suffixes, for example—which have the word they belong to as their context (consider: ing means different things in bring and chopping). Above this stretches a long sequence of ever-larger wholes in which the same sentence can have more or less plausible alternative meanings. First the paragraph, then the chapter, the section, the book, the author’s other works, the author’s life during that period, the totality of the author’s works, the totality of the author’s life, the library in which the works are stored, the culture of which the library is but a single institution, and the sweep of world history in which that culture is but a small part, not to mention the universe itself. Something like the relation between word and sentence repeats itself on a larger scale, though in some cases the connection seems so tenuous as to be nil. Many disputes in literary theory, and perhaps in human studies generally, come down to where to draw the line. Should we interpret Marx’s earlier writings in the context of his later work? Should we interpret politicians by their actions or should we take their promises at ‘face value’?

The holistic relationship between word and sentence reappears between sentence and paragraph as well, although the sentence’s dependence on the paragraph is weaker than the word’s dependence on the sentence. We have already touched on part of the reason—words have no real purpose by themselves. They are incomplete thoughts with no reason for existing aside from their use in the sentences from which they are distilled. But sentences are complete thoughts and so do not need contextualization with equal desperation.

Even so, we know a sentence ripped out of a paragraph, taken ‘out of context’, and treated as a standalone statement can have a completely different meaning. The principles by which this holistic determination works do not seem essentially different from those at play in the dynamic between word and sentence. Generosity is again the guide, though here the consistency and coherence of thought—its unity or logical validity—must play a much larger role than they did for a single sentence. Still, there are sometimes several equally generous readings of a sentence, and sometimes it is the less generous interpretation that is the most plausible. The deciding factor is the context. 

The paragraph is the next level up. The paragraph has a central point—often expressed in what grammarians call the topic sentence, though this need not be explicit—and it is through its relation to the topic that an ambiguous sentence gets its determination. If my topic were writing pens and I said, “Let’s put pigs in the pens!”, a sentence-level interpretation would yield an unsatisfactory result. It would be far more plausible and more generous to override the ‘normal’ atomistic interpretation of the sentence.

All of this is to say, generosity is a multi-faceted principle that operates in many ways at many levels. The overall sense of the paragraph must frequently adjudicate between alternative interpretations of component sentences and/or the relations between them. Sometimes it works the other way around; sometimes an especially cogent atomistic reading of the component sentences determines the point of an ambiguous paragraph. The multi-dimensionality of generosity can result in considerable uncertainty, but there is no question about the influence that a clearly-grasped whole can have on the interpretation of its parts.

How far up does it make sense to go? Clearly the relation of paragraph to chapter and of chapter to book is similar to that of sentence to paragraph, although presumably the dependence of the part to the whole grows weaker the further up the line we go. The more a part can stand on its own, the less likely it is to be ambiguous. The conclusion we can draw from this is that the assumed sense of an entire work adjudicates between alternative or competing meanings of its components, if and when there are such possibilities, all else being equal. In other words, interpretations must equally avoid conflict with facts, historical knowledge, common sense, and the like. After all, it is an empirical matter whether or not Homer was one person or many. No interpretive principle can dictate that whatever has come down to us as a ‘book’ must have a highly unified construction, though the principle does tell us to presuppose truth and unity while solving for meaning to the extent possible or plausible.

How much unity it is reasonable to expect at various levels of organization often boils down to just how much intelligence and control it is appropriate to attribute to the creative energy that brings into being the whole in question. The more a single unifying intelligence controls the creative process—the more unity of purpose, foresight, care, planning, and skill we attribute to the creative energy, and the more power and control that energy is assumed to have over its work—the less we accept explanations of the parts based on accident, circumstance, or external influences. We are entitled to use generosity precisely to the degree that intelligence and power govern the relevant literary whole. At the extreme limit of generosity, we assume the intelligence determines and explains absolutely every detail. At this extreme, the divine inspiration of the poet returns as an homage to the power of the unconscious creative process. Where there is intelligence, generosity is applicable—the more so the greater the intelligence.

In the case of writing, the intelligent or creative force is an author’s mind. But we are unsure that an author’s mind is capable of having a unity of purpose over the span of a lifetime’s output. In the case of a library, even if we assume a single intelligence gathers all the books into a unity and the library’s purpose is constant, that unity has to be comparatively weak compared to the authors who wrote the books. All of which belabors the obvious point that the librarian does not create the books collected.

Even the unity of a single work can be less than crystal clear. In some cases we question whether the work was written by a single author or, as is often the case with ancient texts, whether someone came along later and revised the original. Sometimes the author may deliberately choose not to have a unifying theme, which itself becomes the unifying theme. If nonsense is the point, then nonsensical parts are well suited to make it, and they must be arranged in such a way as to make the nonsensical point with force. But then nonsense becomes the point, which illustrates that even when writers struggle against the notion of unity, it does not really make generosity any less applicable.

Of course, not all literary works are entirely unified, and perhaps it is unclear whether they should be, but it is still true that whatever understanding we have of the whole guides us in the interpretation of the parts. Even when scholarship breaks down a work’s unity, it can do so only by attributing divided authorship to each of its component parts. Within each part, the author’s mind is assumed to reign supreme and that section of text becomes a whole that confers meaning to its own parts. Biblical scholarship is an excellent example of this. A contradiction is felt between the two creation stories of Genesis, which contradiction is thought too great to attribute to a single mind. The book once assumed to be unified and written by a single mind is now viewed as the product of several minds, but—and this is the point—each mind is assumed to have creative control over its own portion of the text. If we did not make this assumption, we could not interpret the text as the conjunction of the thoughts of two distinct minds, one with one viewpoint, the other with another. Thus the disunity gets broken up into two smaller unities, and it is the principle of generosity that tells us when a work is best understood as two works by two authors.

It is here that textual interpretation overlaps with the broadest possible application of the principle of generosity. Unchecked generosity, because it wants to maximize the sense, intelligence, and truth of authors, will want to attribute unity to the largest possible whole. Thus the principle of generosity will urge us to see Homer as a single person—the composer, if possible, or the editor of a collection if that fails. It will urge us to see the Critique of Pure Reason as having been guided by a single vision, even though it was written very poorly and over a long period of time. In order to achieve this unity it is willing and sometimes eager to go far in stretching the meaning of parts so as to eliminate contradiction. Different only is the degree of ‘greatness’ attributed by some interpreters to canonical authors like Plato, Shakespeare, or Joyce.

At the far end of the spectrum is the divine author, the infallible author of sacred scripture, and the need to preserve that infallibility by all sorts of interpretive devices, most of which come down to viewing as metaphorical what would otherwise have been taken literally. (The allegorical interpretation of the Bible was apparently pioneered by Philo in On Abraham and taken into the cannon of Christian hermeneutics by Augustine and Aquinas. It is truly the orthodox view. Fundamentalist atomistic literalism is a curious form of generosity—or piety—in which the interpreter seeks to surrender authority as arbiter of truth.)

There is, however, a crucial difference between the operation of the principle of generosity at this high level and its operation at the level of word and sentence. At the lowest level, the principle of generosity was not an option, but a necessity. When the relevant whole is larger, however, generosity becomes more a matter of choice. Generous-spirited readers will tend to attribute as much intelligence as possible to the author and will assume that whatever disunities they see are caused by their own deficiencies as readers. More critical readers will take superficial disharmonies at face value and attribute them to the author rather than the weakness of their own interpretive abilities. But so construed, the choice is a matter of character and the perceived fit between the details of a particular text and a particular interpretation of it. In such cases there is no hard necessity laid upon us to side with generosity.

Although there is no strict obligation to use the principle of generosity at the higher levels of interpretation, the principle retains even at higher levels a kind of ideal status that strongly compels us to grant more truth to the more generous interpretation, all else being equal. It is very important that this be understood as it has enormous consequences that go far beyond literary criticism.

Imagine we had two ways of doing a jigsaw puzzle, but didn’t know what it was supposed to look like when completed. One way yields a clear picture with a neat rectangular edge, and the other yields a highly irregular border with no discernible image at all. If neither solution to the puzzle has any forced fits of the puzzle pieces, who would deny that the former was the more likely interpretation? The bias is for order, all else being equal (no forced pieces). Although we do not know in advance whether a given work is highly unified or not, we do know in advance that the interpretation that reveals unity will be preferred to the one that does not, all else being equal. Thus the principle of generosity, which was at the lowest level of organization a necessity, reappears at the highest levels as an interpretative ideal—seldom, if ever, achieved, but always desired.

The truest interpretation is, all else being equal, the one that best fits the parts into a coherent whole. To the extent that we cannot make them fit, we cannot understand the work. And while it is true that in some cases the principle cannot be applied so as to illuminate every detail, this is not because there is something wrong with the principle, but rather it is because there is something odd about the work. If we can understand the text at all, the principle applies. Insofar as it does not, we do not have an alternate meaning that might be missed by an over-optimistic use of the principle—we have no meaning at all.

Table of Contents

What do you think?

Questions? Comments? Ideas?

News and Updates:

* This is a short chapter, so the next one is coming soon.

* Download Neal’s Ph.D. thesis on the divided line of Plato’s Republic (PDF) below. I would love to offer you an ebook, but I was working from a typewriter typed document and OCR technology just can’t handle it, apparently.

You can also find it in the “Writing” section.

* Click here to find out how all this subscriber stuff works.

* You can customize which bits of the newsletter you receive by visiting your account at any time. My regular posts will continue to be free.

PART II: Generosity and Shared Belief

Not included in audio

WE HAVE SEEN how the principle of generosity operates in everyday language, and we now stand at the threshold of understanding its greater philosophical implications. In this section I hope to demonstrate how the principle of generosity underlies all communication whatsoever and thereby guarantees the unified, public character of anything worth calling a world.

To do this, I will begin by describing a shift in attitude that occurred in the Western philosophical tradition. I realize it might seem strange to bring up ancient history at this particular moment, but I think it is important to shed light on our own attitudes toward trust, skepticism, and knowledge. Next we will turn to what are known in philosophy as thought experiments, or investigations into the nature of things carried out by the imagination. We will begin by imagining the origin of language and the very first sentence ever spoken, and we will follow this up with Donald Davidson’s famous thought experiment, radical interpretation, which will have us consider what would be required to translate an utterly unknown culture’s language into our own. But first, I have a few things to say about relativism.

7

Politics and Relativism

CONCEPTUAL RELATIVISM, the belief that truth is whatever a particular individual or culture happens to believe, or its linguistic correlate, that language determines thought, is an incoherent doctrine that cannot support its own possibility. Very little in philosophy has been so handily refuted: extreme relativists believe absolutely that absolutism is false.

But to know this is really not enough. We need to understand the cultural hunger to which relativism speaks.

It is important to note we don’t need epistemic equality in order to have political equality and a large measure of tolerance for opposing views. Indeed, that is how the founders of our liberal democracy conceived the matter. It never crossed their minds that all people might be equally right, which is quite different from equal rights. What mattered was only that we learned to tolerate divergent opinions, however wrong or stupid we thought they were.

It took a long time, roughly from 1776 to 1976, for political equality (equality of political rights) to turn into first social equality (equality of income or opportunity) and then epistemic equality (relativism), but it happened. Postmodern pluralism grew out of social and political pressures, particularly in the sixties as people sought more than mere tolerance of previously marginalized ways of thinking for a wide variety of groups. Equality was epistemically extended and taken as an ultimate value. It was a question of demanding respect for what had been held low. Since the various cultural views associated with these groups were tied to specific ways of knowing—the non-verbal, intuitive, emotional, poetic, and so forth—these all had to be equalized. Rival epistemologies led to contradictory conclusions, so it became necessary to speak as if these contradictory ‘truths’ corresponded to alternative realities.

There was and still is real oppression, both political and epistemological, but the blind worship of equality comes at the cost of the distinction between knowledge and opinion. No one on the list of epistemically-privileged groups can fail to get it right, but only, of course, for them. This tinny truth is a lukewarm comfort that pales in comparison to the truth about which one can be wrong. Bluntly put, the product was cheapened to make it readily available. 

Relativism speaks to the demand for an egalitarian society. The question is how to retain its democratic virtues without its nihilistic vice. In the next chapter we will evaluate an ancient alternative to relativism to see what it has to offer.

Table of Contents

What do you think?

Questions? Comments? Ideas?

The very fact you are able to understand the words you're reading right now—that we are all able to communicate with each other using language—means we must share a vast body of beliefs. That’s the central argument of this book:Truth is the condition that makes language possible.

I decided to make this post free. You’re welcome. :)

—Tina

Notes:

* Don’t forget, you can customize which bits of the newsletter you receive by visiting your account at any time—if you don’t want to see the previews of these posts, deselect “Truth and Generosity”. My regular posts will continue to be free.

* Table of contents with chapter links at the end of each post.

* How does all this subscriber stuff work?

* Full audiobook download through Spotify (paid).

6

Social Influences on Semantic Change

WHEN PEOPLE LIE or stretch the truth, they do so because they want to make things appear other than they are, but this does not necessarily involve interfering with language itself. No one rushes to update the dictionary or revise the grammar books just because I said I walked the dog when I, in fact, did not walk the dog. Indeed, if I, a lone individual, made a conscious effort to violate standard language usage to make things appear other than they are, there would be little to no chance of my new usage becoming widespread. Instead, I would probably just get laughed at, as was the case with Reagan when he, in defense of his school lunch program, declared ketchup to be a vegetable. Apparently not even a sitting president can stretch the meaning of vegetable to include a condiment consisting largely of high fructose corn syrup. But whenever the passion to distort the truth is shared by a great number of people, especially those who have great influence, changes to language are possible, and such doublespeak can be used as a vehicle of distortion. Words are then put under a truth stress and their meanings change accordingly. But the process of semantic shift takes time, perhaps enough time for the untruth to accomplish whatever it was intended to do.

To give a concrete illustration of this process, we start with a relatively recent semantic shift for which there is abundant firsthand knowledge. Unfortunately it concerns a symbol rather than a word, but in this example the procedure for semantic adjustment clearly stems from untruth, which is why I think it is a good place to begin.

In the late sixties and early seventies, college professors began grading with Bs work that would previously have received a C. As a result, grades changed meaning. A B or even a B-, which had previously been a mildly honorific grade, came to mean merely acceptable or average, and a C, which previously meant acceptable or average, came to mean below average. 

It didn’t take long for this Vietnam-era grade inflation, which was well-documented at the time, to become well-known to those with an interest in these matters. A few attributed the inflation to students simply performing better in school during that time than at any other time in the past fifty years. Others saw it, more plausibly, as part of the trend toward egalitarian ideals in education. But most believed it to be a response to military draft requirements, which at the time allowed deferments for college students. The unpopular Vietnam War draft, it was thought, put pressure on professors to avoid flunking students out of college—grades, after all, had become a matter of life and death.

Whatever the motivation behind the change, the fact is a very large number of professors chose to tell lies—or untruths, exaggerations, falsehoods, call it what you will—but these were only effective for a short time. Once lying becomes normal it ceases to be effective, ceases even to be a lie. The language of untruth morphs into a new vocabulary for truth. Thus the new B soon became translated as the old C, and things went on much as they had been before. In the new language, students are stung or pleased by a B- just as they were once stung or pleased by a C. Similarly, faculty who once felt queasy about giving a B to an average student now do so without misgivings. Faculty who initially refused to go along with the inflation are now forced to do so or they unfairly penalize their own good students by speaking an archaic language. The language has shifted, despite the intentions of its speakers, and now anyone who wishes to be understood must adopt it.

Political motivations for semantic change are abundant. Politics is, after all, an arena in which large numbers of people share the same strong passion. From this inevitably stems a struggle for the power to influence others and to control the means of communication, foremost among which is language itself. Socially-motivated semantic change can be as significant as a desire to change society’s valuation of women or it can be as trivial as a preference for meat over vegetables—a preference shared, perhaps, by a certain former President of the United States.

The word meat is an interesting case. In Old English, mete, maet, or meate simply meant food. This usage can be traced back to written texts dating from 900, and can still be found in odd corners of the language as late as 1900. The current meaning of the word is only the third given meaning in the Oxford English Dictionary and the earliest example of that usage is from 1460. Presumably somewhere in the preceding 600 years the meaning of the word was narrowed from food to the particular kind of food we now call meat. We can only speculate on how and why this happened, but since it was certainly not dictated by some central linguistic authority, there must have been first users who, given the then-current meaning of the word, spoke in some form of falsehood. Perhaps their preference for meat was such that upon being served a plateful of vegetables, they said, “That’s not food (mete).” Or maybe when the roasted animal flesh was being served, they said, “Ah, here’s the real meal (mete).” This supposition is strengthened by the still-currant usage of the word in phrases such as the meat of it or essence of the matter. The meat was and still is considered by many to be the real thing, the main course, the essence of the meal, the part really worth eating, or even the only edible part, as in the meat of a walnut or coconut.

Whatever the case may be, the widespread tendency to value meat above all ended in the appropriation of the entire category of food (mete), thereby demoting vegetables, previously called greenmeates. If there had been an organized force of vegetarians at the time, it might have vigorously protested this abuse of language, but to no avail, for the meat lovers proved more influential. They controlled public language, and for a while this might have given the meat partisans great satisfaction. In the long run, however, their lie or exaggeration was destroyed by the principle of generosity which altered the meaning of the word mete so as to turn their boast into the harmless tautology that animal flesh is meat.

The more controversial man has a similar history. In the oldest preserved English usage, dating from around 825, its primary meaning is simply the genderless human being. This was close to the impersonal German Mann and had about the same meaning as the current one in everyone. A woman was generally called a wifman, (as in wife) or female man, and a man was usually a werman (as in werewolf) or male man. A secondary meaning of the word was adult male human being in both Old English and all other Teutonic languages, but only in English did this secondary usage supersede the primary meaning. Just as the narrowing of mete represented the point of view of the meat eaters, so for man it surely represents a tendency towards male-dominance among those who controlled language, presumably the men. The analogy rolls toward its ironic dénouement: just as it did in the history of mete, the principle of generosity foiled the effort to promote untruth. Over hundreds of years the outrageous claim that only men were human was whittled down to the harmless tautology that only men are males.

The examples we’ve discussed have all been of semantic change brought about by whichever group had the power to spread untruths, exaggerations, or outright lies. The same process happens when there is no intention to deceive. The word robin, for example, in pre-American English once referred to a certain type of red-breasted flycatcher (F. Muscicapidae) common in England. The orange-breasted American thrush (F. Turdidae) had a superficial resemblance to this bird, and this led early settlers to incorrectly call the American thrush a robin, which it most certainly was not, given the then-current meaning of the word, though now it is perfectly correct in American English.

Consider the word dilapidated. In current standard English it has the ‘literal’ meaning: To be in a state of disrepair. But the etymology reveals the word once referred only to stone buildings. Through a process of generalization the word came to have its current meaning. Presumably whoever first used the word dilapidate to refer to a wooden structure was something of a poet, or had no better word to describe a wooden house that was falling apart. The misuse of the word turned out to be useful, and it stuck.

Each of these examples is a case of a word being used to do a job it was not designed to do. Sometimes the misuse is done on purpose. Sometimes not. In either case there follows an adjustment of meaning that forces the word to fit the sentence under the presumption of the sentence’s truth. Much of this happens when we are faced with new circumstances and an old vocabulary, and that in turn happens whenever the human mind moves into a new region, whether that region is geographic (America) or metaphorical (the brave new world of wooden houses falling apart, green blackboards, subatomic physics or religious experience). In all such cases we are under pressure to misuse words, the meanings of which are constantly racing to catch up with our abuses by adjusting themselves to make our falsehoods true, to make the unbelievable believable

Table of Contents

What do you think?

* Can you think of instances where you were “under pressure to misuse words”?

* Can you come up with examples of word meanings that have adjusted themselves to make our falsehoods true?

Questions? Feel free to comment!

The very fact you are able to understand the words you're reading right now—that we are all able to communicate with each other using language—means we must share a vast body of beliefs. That’s the central argument of this book:Truth is the condition that makes language possible.

To take the poll, click on your answer choice. Your answer is completely anonymous, not even I will know. This is just for fun.

Notes:

* Don’t forget, you can customize which bits of the newsletter you receive by visiting your account at any time—if you don’t want to see the previews of these posts, deselect “Truth and Generosity”. My regular posts will continue to be free.

* Take the last poll: Is reality beyond linguistic expression?

* Table of contents with chapter links at the end of each post.

* How does all this subscriber stuff work?

* Full audiobook download through Spotify (paid).

If you prefer paperbacks or ebooks but don’t want to miss out on the discussion on future paid posts, just let me know you’re reading along the good old fashioned way (I trust you) and I’ll get you past the paywall. :)

4

What Language is Not

Language is not a Code

TO SAY LANGUAGE is not a code is to deny any strong notion of literal meaning. Of course, everything depends on what is meant by literal. The word itself is an amusing example of the problem. Etymologically it means something like in the letters of, as if the meaning of a sentence could somehow be physically analyzed and found in the marks on the page.

It turns out the word literal is itself a dead metaphor, the meaning of which we try to express through words such as explicit, non-figurative, or prosaic, the opposite of which we call implicit, figurative, or poetic. These notions overlap, all are vague, and several are metaphorical. Figurative, for example, compares non-literal language to figures or images, and implicit compares that which is indicated but not said to the condition of being hidden within the folds of something, presumably clothing. It is tempting to cut through all this by defining the literal as that which needs no interpretation, that simply gives of itself, or without further ado, the meaning it intends to convey. But then everything depends on what is meant by interpretation.

No part of language is so straightforwardly ‘on the page’ that its comprehension does not require knowledge of the meanings of signs and the rules for their combination. Even so, it’s not enough to consult a dictionary and grammar book. Understanding ordinary, natural language requires more than merely following the conventions that constitute normal grammar and semantics; it requires the principle of generosity, as we’ll soon see.

Coded communication like Morse code, on the other hand, is an invention that operates in a mechanical, rule-governed way, solely through consultation with a code book. Of course, there are similarities between coded communication and natural language—both make use of external symbols which have little to no intuitive connection to the things these signify—but it would be a mistake to say that natural language is a code or even code-like. Natural languages are bursting with figurative speech, but codes are, strictly speaking, literal representations of their meaning. In Morse code, for example, three dots on a page are replaced by the letter s in strict accordance with the rule. So also for the other marks on the page, and when we are done with these replacements, we have, without further ado, the message. Except in the case of telegraphic error, there is no need for any further interpretation to get from Morse to English—but much more will be needed to get from English to the final meaning.

To the extent that natural language is ambiguous, there is, strictly speaking, no message sent. What is sent is instead a sort of half-formed, raw, and rather complicated collection of possibilities that need to be sorted through and ranked in order of plausibility.

Dictionaries may look like code books, but they are far from sufficient for understanding meaning. That is because all natural language is highly ambiguous and therefore requires interpretation even in its most commonplace, unimaginative employment. Indeed, it is in its commonplace employment that interpretation is most needed. Only highly cultivated, artful and perhaps artificial speech makes any pretense to unambiguous language, and it is only such language that suggests (falsely, as we shall see) any serious resemblance to code.

The figurative nature of language is probably most clear in slang—the word cool, for example, or awesome—but this applies equally well to high-toned language (a brilliant scholar) as to mundane utterances (a colossal bore). It turns out that many of what are considered standard, literal meanings of a word are just metaphors that have become so familiar that we no longer think of them as such.  These are what we might call dead metaphors. For example:

She had a very fine understanding of the matter.

Sally is a neat housekeeper.

Dead metaphors are all but invisible. Etymology reveals a history of words taking on metaphorical usage and gaining currency until they finally become standard. When they become thoroughly established, we no longer see them as metaphors. 

A closer look reveals that even the simplest bits of language turn out to be highly figurative, which means natural language cannot be a code or even like a code, for it includes very little, if anything, that is literally literal. Meaning is almost never on the page or in the letters. It is in their interpretation guided by the principle of generosity.

Language is Not Invented

Like the fire and the wheel, it is sometimes assumed that language is invented. Presumably this is because we think of language as code-like, and codes are invented and sometimes even named for their inventor, like Morse. But since, as we’ve seen, language is not a code, the analogy is suspect. It only takes a little consideration to realize that it is also not an invention, at least not by any standard sense of the word.

Codes and writing symbolize speech, but speech symbolizes thought. The connection of thought to speech is clearly very different and far less direct than the connection between one symbol system to another. It makes little sense to suppose that what is true of codes is also true of speech. One might object that coded and written language can to some degree take on a life of their own and enter into a more direct relationship to thought, which would make them more like natural languages. This may be true, but to whatever extent this happens, it happens through the mysterious processes of evolving usage, not the deliberate processes that characterize invention.

Neither code nor writing could have come into existence had not the spoken language arrived first. Turning thought into speech requires something altogether different from replacing one symbol system with another. Of course, we can’t go back in time to witness the birth of language, its origin can never be known, but whenever anything turns up that is both language-like and clearly an invention, language is already there. Even Esperanto, which is often called an invented language, on closer inspection turns out to be an elaborate code, which is to say it involves translating pre-existing symbols into other symbols. 

Not only is there no single example of invented language, but it is hard to imagine how there ever could have been such a thing. Invention presupposes deliberate, self-conscious, and fairly sophisticated thought processes that cannot with real confidence be attributed to creatures who do not speak. We don’t say that beavers invent their dams or that birds invent their calls. We don’t even say chimps invent the tools with which they pull ants out of holes. Invention, at least in the usual sense of the word, seems to presuppose speech; if so, the invention of speech is an incoherent idea.

Language is Not a Convention

If language is neither invention nor code, then it is not likely to rest on consciously-established conventions either. It is difficult to see how language could have arisen from the act of assigning meanings to words since that would require a high level of communication already in place. In other words, a community that is capable of agreeing to conventions must already have a common language. Not only must we already be able to speak, but we must be able to speak about speaking, which amounts to a rather high level of linguistic self-awareness.

In the old speculations on this matter, the anti-conventionalist view was associated with those who considered the relationship between sign and signified in some sense natural. In their view, the first words were expressive song (Jespersen), instinctual interjections (Bleek, Noire), or onomatopoeia, the imitation of animal sounds or sounds that were supposed to be typical of the thing being described (Heyse, Max Mueller). While there may be some truth in these suggestions, my opposition to the conventionalist thesis does not include these.

Whether language originated with nature calls or the Neanderthal equivalent of the Academie Francaise, it seems clear that meaning evolves through usage that is, on the whole, under the conscious control of no one. The current meanings of words have almost nothing to do with the establishment of conventions. We might assume word’s meanings are first learned and then used in combination with one another to make statements which may or may not be true. This is an atomistic view in which the statement’s parts, its words, come first, and the truth of the whole sentence, if it comes at all, limps along behind. But Davidson’s treatment of linguistic irregularity—the whole range of phenomena characterized as malapropism (flamingo dancing, for example), idiolect, and other forms of misuse—suggests language operates through a holistic process that holds truth constant and solves for meaning. In other words, he argues that the truth of a sentence comes first and the meanings of its words are adjusted to make this truth possible. I will go into this in more detail later, but for now it is enough to note that it is not convention that creates the meanings of words, but our beliefs about the world combined with the principle of generosity.

Suppose an auto mechanic from the rural south currently living in New England tells me that in order to deal with my car problem, I’ll need an auto holler. I look at him in genuine confusion, my mind half-forming the question: Can he really think yelling at the car will do any good? But surely not. He goes on to tell me that he’ll call the auto holler from the neighboring town and send it down to Northampton to get my car. He adds that the driver should get here in just a few hours. By this point I realize he is using the word-sound holler as I would use the word-sound hauler, and from that point on everything proceeds smoothly.

There is little mystery here. This is an obvious case of ‘mispronunciation’ or differences in regional dialect, and we figure out the meaning of the strange sign from its context. The initial confusion does not get straightened out by anyone’s explaining to me the conventions of Southern dialect. Instead I presupposed the truth of the mechanic’s speech—in other words, I assumed the mechanic was quite aware that yelling at my car would not solve a thing—and from this presupposition of truth I realized that the meaning of the word-sound holler in his language was hauler in mine. I presupposed the truth and solved for meaning. From now on I will go directly to that meaning whenever speaking to this particular person or to people who speak like him. The point is, the truth of the sentence came first, the word and its meaning came second. I can be sure then that for me, the word did not get its meaning by convention, but by generous interpretation of its usage in accordance with my beliefs about the world.

Suppose that instead of a dialect we were dealing with an idiolect, a deviant usage that is not common to any group, but pertains only to a single person. An example of this might be a philosophy student who keeps calling arguments ad nauseam when he really means ad hominem, or someone who consistently misspells perceive, or someone who pronounces Vermont as if it were Vairmont (I know such a person). It is clear that in such cases we do exactly what we did in the previous one; we use the principle of generosity to establish translations from the idiolect into the public language. Again, communication is not made possible through semantic agreement. Instead it becomes clear that, for whatever reason, the speaker uses language in a different way, and the audience simply figures out what is meant by assuming that the speaker speaks the truth.

In idiolect the deviant usage is consistent, and this helps us see it as a semi-private language. But suppose it is the kind of error that occurs on a single occasion, as when, after a few drinks and, perhaps, a discussion of swamps, I say that I am going to walk the bog. I could be joking, in which case the misuse is intentional, or it could have been a simple slip of the tongue. In both cases the audience does the same thing. They translate bog into dog, which is to say they treat my utterance as a slightly foreign language and translate it into their own, guided by the principle of generosity.

For dialect, idiolect, and slips of the tongue, the deviant word meanings are established not by a table of conventions, but through usage and generosity. In these cases meaning is plastic and takes its shape by conforming to the contours of a presumed shared reality. 

We have seen how most, if not all, established usage was once deviant usage that has gone through a process of normalization. I suggested above that this process is governed by the principle of generosity, and that virtually all words have acquired their meanings in ways similar to the discovery of the meaning of holler in the example. I touched on this point in my discussion of etymology, but this is a topic which deserves a chapter of its own.

Table of Contents

What do YOU think?

* Do you agree that “meaning evolves through usage that is, on the whole, under the conscious control of no one”?

* Does invention presuppose language?

Questions? Ideas? Feel free to comment!

Philosophy and Fiction is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Thanks for reading and sharing your thoughts.

We appreciate your generosity!

Neal, Tina, & Geordie Bear

It doesn't matter whether you're conservative or liberal, religious or skeptical, the very fact you are able to understand the words you're reading right now—that we are all able to communicate with each other using language—means we must share a vast body of beliefs.

While language may shape the way we think about the world to some small extent, it makes much more sense to say truth shapes language to a very large extent.

That’s the central argument of this book:Truth is the condition that makes language possible.

To take the poll, click on your answer choice. Your answer is completely anonymous, not even I will know. This is just for fun.

* Poll: Given our current political divides, would you say we live in different realities?

* How does all this subscriber stuff work?

* Full audiobook download through Spotify (paid).

* Table of contents with chapter links at the end of each post.

We appreciate your generosity!

Neal (Santa Claus), Tina, & Geordie Bear

3

The Poetry of Ordinary Language

TO DEMAND THE LITERAL TRUTH from poetry and obscure sacred texts—a demand we might call one of the defining characteristics of Western philosophy—is to ask for something that was never promised and cannot be delivered.

In Eastern philosophy, on the other hand, it is common to view language as figurative. Take, for instance, the Hindu and Buddhist perspectives on the metaphysics of maya, the notion that reality dissolves the boundaries that hold things apart. Because language depends on the separations that define maya—subject from predicate, noun from verb, species from genus, one word from another, and so forth—it is assumed that reality is beyond linguistic expression. But of course this does not mean that one ceases to talk about it. It merely changes the way we view such talk. In effect, it increases the importance of poetry in our conception of speech and knowledge. It is simply assumed that all talk is essentially poetry, whether or not the speaker realizes it. It is only through poetry—words that when taken literally are untrue—that we have any way at all to speak of reality.

What does it mean to say that God has the head of an elephant and is the son of Shiva and Shakti? It’s hard to say, even if we worship God this way. What does it mean to say that I am drinking the blood and eating the flesh of Jesus? One can cherish the ritual and still have no idea what it means. Does the notion that God has the head of an elephant contradict the idea that God had a son named Jesus? What about the idea that Jesus is both God and the son of God? Only the illusion of literal truth makes us wonder whether there is a contradiction here. Once we free ourselves from that, the words and images become supple and malleable, mere riddles in need of interpretation.

A dramatic shift must occur in one’s attitude toward logical contradiction. The principle of non-contradiction applies straightforwardly to literal speech. In it, A and not-A cannot both be, as we say, ‘literally’ true. Indeed, the acceptance of non-contradiction as the measuring rod of truth is likely one of the defining qualities of literalness. But in poetry, the principle of non-contradiction either does not apply or applies only indirectly. In one line of (not very good) poetry, I can describe the sun as a blazing bird, and in the next I can call it the angel of the daytime sky. No one but a pedant will bat an eyelash. In the same way, the Hindu can say that God is Shiva and Shakti and Krishna and Ganesh and any of a thousand other gods. He can with little trouble add Jesus, Buddha, Gandhi, his guru, or the girl next door to the standard pantheon precisely because he understands all these images of God as merely that—images, metaphor, poetry—which, although they are to be passionately embraced, cannot be taken literally. Because they are images, all mirror an aspect of the godhead; they are all ‘true’. But for the same reason they are also ‘false’ and need to be taken with a grain of salt. They make sensual and imaginable what is in itself none of these things. Because of this, images are both precious and not to be taken seriously.

Many religions believe taking a picture of God literally is idol worship. There is a Hindu prayer that apologizes to the universal, timeless, and shapeless God, asking for forgiveness on the grounds that it was only by giving God some definite form that it was possible to worship him at all. But the Hindu, precisely because he has so many idols, has none, whereas literalists with only one well-guarded image, are, at least from the Hindu perspective, the true idolaters.

What is of interest here is not the metaphysics of any particular religion or any sort of theology, but to point out what happens when we think of language as poetry. Insofar as metaphors do not pretend to be literal depictions, alternative metaphors cannot contradict one another. Their meaning is far too obscure for simple logic.

How sharp is the line between poetry and ordinary speech? More blurry than we commonly realize. We need only consider how much of ordinary speech is slang or idiom. A great deal of it, if not most of it. And surely most slang and idiom, whether evocative or dull, is essentially metaphorical. Consider for instance the words: cool, neat, lousy, and stinks. Or these ubiquitous phrases: passed away, home free, sign up. Anyone who has attempted a word-for-word translation into another language will quickly come to realize how figurative language really is. And because virtually the whole of ordinary language either is or was figurative, virtually the whole of it either needs or once needed the principle of generosity. But to really make this case, we will first need to dispel some common misconceptions about language.

Table of Contents

What do YOU think?

* Do you agree that “virtually the whole of ordinary language either is or was figurative”? Why/why not?

* Is reality beyond linguistic expression?

Questions? Ideas? Feel free to comment!

Chapter 2: Violations of the Principle of Generosity

Given our current political divides, would you say we live in different realities?

There are invisible things the necessity of which is most poignantly known by their absence. Air is one of them. The principle of generosity is another...

Click here to get Truth and Generosity in paperback!

Click here to find out how all this subscriber stuff works.

You can customize which topics within this newsletter you receive by visiting your account at any time. My regular posts will continue to be free.

We appreciate your support! Thank you!

Take the last poll!

PART I

The Principle of Generosity

Not included in audio.

In this section, we will see how the principle of generosity operates in ordinary, everyday language and how its absence affects our ability to communicate. We compare everyday language to poetry to show how figurative even the most ordinary speech actually is, despite common misconceptions about language. Finally, we will show how generosity drives semantic change, thereby preserving our ability not only to speak the truth, but to say anything at all. 

1

The Principle of Generosity

Included in audio!

IF WHILE WALKING DOWN a city street we overheard someone say, “Fiona put the pig in the pen,” we could take the words in what might seem their ‘normal’ sense, or we could interpret the remark to mean she put the pig in a writing instrument. Properly speaking, there is no single normal sense of the word pen. Nor is there any meaning that is obviously the most common. Nothing about its pronunciation or spelling gives any indication of which is meant. The word is straightforwardly equivocal; in other words, it has at least two perfectly normal and equally proper dictionary definitions. And yet, something like the decision to take the word as meaning enclosure for animals is made, and it happens nearly as soon as the sentence is heard. Maybe even before; the word pig may automatically dispose us to consider only the pertinent meaning of pen.

Of course, in this sentence it would be perverse to assume the word meant fountain pen—a joke, at best, at worst an annoying act of small-mindedness. But why? What makes one meaning immediately obvious and the other a tedious joke?

It might be thought that the context makes the meaning clear. But the context alone forces nothing, for it, too, must be interpreted. Of course, there are different kinds of context that can affect meaning (which we will discuss later), but to understand, Fiona put the pig in the pen, the sentence is all we need to choose the meaning enclosure for animals over writing instrument. When reading the sentence, the pig is in the pen, it would be odd if it even occurs to you to wonder whether the word pen means writing instrument.

Because freestanding words in natural language are often too ambiguous to be useful, it is the sentence itself that tells us how to take the word’s meaning. This is what we might call an organic or holistic relationship in which we interpret words in such a way so as to yield a whole that, as we say, makes sense. For declarative sentences, this means interpreting the words in such a way that the sentence is, is likely to be, or at least might possibly be true according to our beliefs at the time. This brings us to the connection between truth and meaning on which everything here depends.

It comes to this: Given our beliefs about pigs and pens, Fiona could not possibly have put a pig inside a writing instrument, so pen must be taken to mean an enclosure for animals. Pigs don’t fit inside writing instruments—such an assumption is so mundane we are probably not even aware of making it. In interpreting ambiguous sentences, all else being equal, we choose the possible over the impossible, and the likely over the unlikely—that is to say, we choose what is most likely to be true. The command to interpret the sentence so that it ‘makes sense’ is just the command to interpret it so that it is most likely to be true. Thus we come quickly to the principle of generosity, which tells us that, all else being equal, we attribute to the words of any sentence the meanings that are most likely to make the sentence true.

We can, of course, imagine a different scenario in which the correct interpretation of the sentence would be far from automatic, even difficult. Suppose we are talking about fountain pens—our favorite types, their superiority to ballpoints, and so on—and you suddenly say, “Fiona put the pig in the pen.”

Then I, thinking we are still on the same subject, ask, “What?”

And you, distractedly, reply, “She put the pig in the pen, but I forgot to tell her about the feed.”

And I ask, “What pig? What are you talking about?”

And you answer, “The pig at my farm. I just remembered I had Fiona put it in the pen but I forgot to tell her where the feed was.”

In this conversation, different contexts and interpretive principles conflict with one another. Initially we are guided by what we might call the principle of coherence. We assume what is being said is relevant to what we were talking about just then, which leads us to the bizarre idea of Fiona putting a pig in a writing instrument. But this interpretation conflicts with the principle of generosity. All else is not equal, and so we are confused and the conversation teeters at the edge of breakdown. It requires a bit of effort to clear things up. You tell me, in effect, that the principle of coherence is not applicable here and that the principle of generosity should be allowed to determine the matter. Communication is then restored, although I may be irked or amused by your dreamy way of talking. 

Of course, we could imagine a scenario in which pen in the very same sentence really does mean writing instrument. Suppose we are playing a game called Find the Pig which involves hiding teeny tiny plastic pigs and telling others where to find them. In that case, the pen in Fiona put the pig in the pen would clearly mean writing instrument. But this does not gainsay the general point about the principle of generosity. It simply clarifies how, in other circumstances, the same principle gives different results.

The principle of generosity is not the only principle that guides interpretation, but it is an absolutely essential component in understanding equivocal language. This does not mean that the most generous interpretation is always correct. It only means that all else being equal, and there is usually much else, the generous interpretation is the most likely.

Although the principle of generosity can be explicitly formulated and we can consciously adopt it and make it something of a moral obligation to extend it as far as possible, perhaps as an act of courtesy, similar to the way some anthropologists interpret foreign cultures and some scholars interpret books, my point is not that there is any conscious act of kindness here, nor that we have any obligation to be generous. The generosity I am talking about is not optional. Consciously or unconsciously, whether we ought to employ it or not, we simply must and do use it in order to understand language. Without it, communication through the equivocal medium of natural language would not be possible.

If this seems an exaggeration, perhaps that is because we do not normally notice how much equivocation characterizes all natural languages. In our example we only discussed two possible meanings of the word pen. But consider the following:

Let’s send the criminals to the pen.

The pen is mightier than the sword.

I hope to pen a letter soon.

Consider the meanings of pig in the following sentences:

The pigs came down on us in a fury and broke up our peaceful demonstration.

Put the pig back in the furnace, heat it to 900 degrees, and mix it with more raw ore.

He was a pig at dinner.

Consider put in the following sentences:

He put a false interpretation on events.

He put his house in order.

She put the shot high in the air.

Put your money where your mouth is.

They put the time at 5 PM.

She put it rather bluntly.

They put the question to a vote.

He put it to music.

Consider in in the following sentences:

He was hit in the head. 

She’ll be ready in ten minutes.

They were put in charge.

He paid in cash.

Not in my opinion.

You’re in for a big surprise.

Virtually every word in the pig-pen sentence is equivocal. The possible combinations and permutations of even the most standard meanings of these few words produce a bewildering minimum of 500 interpretations of the sentence, each of which is equally legitimate so far as a strict decoding process goes. Thus it seems clear that a constant unconscious use of the principle of generosity is at work in almost every bit of natural language communication.

This conclusion can be avoided only by finding another principle of selection. I can think of only one contender, which I will call the principle of normalcy. Such a principle would say that there is a default or normal meaning of any sign, where normal would be taken to mean most frequently employed. This principle is especially appealing when, for example, correctness requires us to choose what seems to be the ordinary meaning of put over its rarely used, highly specialized meaning in Olympic athletics.

But it is obvious that we do not go by the rule of normalcy alone. If that rule led to a blatantly untrue or nonsensical meaning, and if there were less common meanings of the words that did not, we would automatically search among the latter for a better interpretation.

There are other, perhaps equally serious, problems with the principle of normalcy. Often there is no perceived difference in normalcy between two meanings or it is unclear what the normal usage is. Is it the most frequent? In the case of the word in, we might think its most normal meaning is spatial or locational—after all, this is the meaning that comes first in both the Oxford English and Merriam-Webster dictionaries—though very likely its most frequent meaning is metaphorical. But then we have to figure out which metaphorical meaning is the most normal or common. I do not doubt that there is a principle of normalcy that plays an important role in the interpretation of equivocal language, but it clearly operates only in conjunction with the principle of generosity. One of the conclusions we can draw from this, not one stressed by Davidson, might seem implausible at first glance: If the principle of generosity works here as claimed, then it looks as if the principle operates constantly and unconsciously in all linguistic communication.

I have spoken of what goes on in interpretation as if the process were an elaborate cognitive interpretive mechanism, but this must itself be taken in something of a generous spirit. Conceivably the principle operates more as an attitude than a method, or perhaps there are other mechanisms. It makes no difference to me; I find it useful to speak as if the principle were a matter of cognition, even if it is not literally true.

What I have said thus far is that ordinary language is highly poetic, metaphorical, idiomatic, figurative, and eccentric in a wide variety of ways, such that even the simplest words turn out to be equivocal and in need of interpretation. While there may be other principles at play, none of these alone is sufficient in interpreting natural language; the principle of generosity is always required. 

Philosophy and Fiction is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

Contents

Some of these links won’t work if the chapter hasn’t been published yet.

Introduction

Chapter 1: The Principle of Generosity

Chapter 2: Violations of the Principle of Generosity

Chapter 3: The Poetry of Ordinary Language

Chapter 4: What Language is Not

Chapter 5: Etymology and Truth

Chapter 6: Social Influences on Semantic Change

Chapter 7: Politics and Relativism

Chapter 8: Trust and Doubt

Chapter 9: The Origin of Language

Chapter 10: Radical Interpretation

Chapter 11: How We Recognize Language as Language

Chapter 12: The Body of Truth

Chapter 13: The Heart of Truth

Chapter 14: Generosity Beyond the Sentence

Chapter 15: The Interpretive Ideal

Chapter 16: Interpreting the World through Generosity

Bonus Chapter: The Socratic Dialectic and Generosity

Front and Back Matter

What do you think?

Most of this book was written over 20 years ago, but since then, interesting language learning research is being carried out today. Our current AI language models operate on huge data sets which far exceed what a child encounters, require feedback and correction from humans, and yet they ‘hallucinate’ and produce inhuman errors. This would seem to indicate something more is needed, that innate conceptual knowledge or something like ‘common sense’ underlies human language acquisition. Is it possible to create AI that truly communicates and understands language the way we do?

Do you have questions about what you’ve just read?

Ideas you’d like to talk about?

Leave a comment!

Introduction to Truth and Generosity: How Truth Makes Language Possible