Capítulo 4
The Hidden Structure of Verbs
Like Alice stumbling into Wonderland, exploring the seemingly mundane problem of how children learn verbs reveals a hidden microcosm of basic human ideas and their connections. Children's language acquisition represents an astonishing feat of induction: from finite samples of speech, they extract infinite generative rules. Within just three years, without formal instruction, children develop vocabularies of thousands of words and command complex grammatical structures.
Their errors (like "All animals are wake-upped") reveal they're not merely imitating but analyzing language systematically. This represents a profound solution to what philosophers call the "induction problem"-deriving universal principles from limited examples.
Verbs present a particularly fascinating puzzle. Beyond the basic distinction between intransitive verbs (like "snore") and transitive verbs (like "sprain"), English has verbs requiring oblique objects, sentence complements, and other patterns, with thousands of verbs falling into dozens of classes. A verb determines not just word order but meaning-contrast "Barbara caused an injury" with "Barbara sustained an injury."
Children learning language face a puzzling challenge with verb constructions. When encountering verbs like "load," they discover these verbs can appear in multiple constructions (like "load hay into the wagon" and "load the wagon with hay"). This pattern appears with other verbs like "spray," "splash," and "rub," suggesting a rule where content-locative constructions can alternate with container-locative constructions.
Yet many verbs resist this alternation. "Pour" works in "pour water into the glass" but not "pour the glass with water," while "fill" works in "fill the glass with water" but not "fill water into the glass." The paradox deepens because children do generalize, creating novel constructions like "Can I fill some salt into the bear?" They aren't simply memorizing verb-construction pairings, nor are they consistently corrected when they overgeneralize.
The breakthrough in solving this paradox came when linguists reconceived the problem not as phrase manipulation but as a conceptual gestalt shift. The two constructions represent fundamentally different event construals: the content-locative ("load hay onto the wagon") means "A causes B to go to C," while the container-locative ("load the wagon with hay") means "A causes C to change state by means of causing B to go to C."
This elegant solution reveals how deeply perspective-taking is embedded in language-even in describing simple physical actions, we can frame events in different ways. This framing difference explains the "holism effect": when something is expressed as a direct object, it's interpreted as wholly affected. The pattern reveals how English treats changing entities similarly to moving entities, conceptualizing states as locations in a metaphorical state-space-demonstrating the pervasive role of metaphor in everyday language.
Capítulo 5
The Language of Thought
When speakers instinctively avoid certain verb constructions or wince at others' usage, they're demonstrating sensitivity to subtle semantic distinctions that are never explicitly taught. This sensitivity suggests people represent verbs in memory such that those with similar syntactic behaviors share overlapping definitions in an abstract language of thought.
This mental language must be highly abstract to group verbs by their conceptual structure rather than surface similarities. For example, "hand," "carry," and "bring" might look similar in the physical world, but syntactically they behave differently because they represent different conceptual categories. Similarly, "baking a cake" is conceptually grouped with "building a house" (creation verbs) rather than with "warming a cake" (state-change verbs).
While specific constructions aren't universal across languages, similar patterns repeatedly emerge worldwide. Locative alternations appear in languages from German to Japanese to Bantu languages; dative constructions exist on every continent; and causative alternations number in hundreds of languages. When these constructions appear, they typically follow the same conceptual scripts-like the "holism effect" in container-locatives.
Certain fundamental concepts consistently emerge across languages: basic entities (event, state, thing, path), relationships (acting, going, being, having), taxonomies of entities, spatial concepts, time, causation, and goals. These form the major words in our language of thought-a conceptual scaffolding with hooks for hanging images, sounds, emotions, and other contents of consciousness.
Brain studies support this distinction between core grammatical concepts and richer meanings. Patients with certain brain damage lose ability to distinguish similar verbs (like "drip" from "pour") while retaining sensitivity to their grammatical behavior, while others show the opposite pattern.
The fundamental concepts that structure our language-causation, force, time, and substance-aren't universal logical necessities but evolved mental tools that sometimes fail us in modern contexts. Several examples illustrate these limitations:
Having and benefiting: Though we conceptualize well-being as possession ("rich is better"), research confirms that beyond basic affluence, additional wealth brings little additional happiness.
Having and knowing: The "conduit metaphor" treats knowledge as objects we possess and transmit, but cognitive science shows this misleads us. Memory doesn't store information like objects on shelves but actively reconstructs it, often inaccurately.
Time: Language packages time into discrete events rather than continuous flow, missing the notion of time as a measurable commodity-perhaps explaining cultural clashes between hurried postindustrial societies and more "lackadaisical" approaches to time.
Causality: Our prototype of causation involves direct, intentional action, but real-world causality often involves complex chains of events that challenge legal concepts of responsibility, as in cases where someone indirectly causes harm.
Capítulo 6
The Metaphor Metaphor
George Lakoff, former student of Chomsky and founder of Cognitive Linguistics, represents the strongest advocacy for what might be called the "metaphor metaphor." In his view, metaphor isn't mere linguistic ornament but essential to thought itself: "Our ordinary conceptual system is fundamentally metaphorical in nature."
Lakoff makes radical claims based on this premise. Since we think in metaphors grounded in physical experience rather than logical formulas, he argues the entire Western philosophical tradition is misconceived. He rejects objective truth, seeing only competing metaphors that serve different purposes. He reframes Western philosophy as a succession of conceptual metaphors, mathematics as bodily-based rather than eternal truth, and political ideologies as rival versions of the SOCIETY IS A FAMILY metaphor.
But people can transcend metaphors in their language rather than being trapped by them. Experimental psychologists like Keysar and Glucksberg have shown that people don't access the underlying conceptual metaphor when processing conventional metaphorical expressions. Only when metaphors are fresh and unconventional do readers activate the underlying conceptual mapping.
We can question, discount, and analyze metaphors-even joke about them, as in Steven Wright asking "If all the world's a stage, where is the audience sitting?" This ability to analyze metaphors reveals that we must possess an underlying medium of thought more abstract than the metaphors themselves. When reasoning with metaphors like LOVE IS A JOURNEY, we focus on relevant aspects (shared destinations, progress) while ignoring irrelevant ones (packing luggage, gas stations).
Beneath metaphors lie abstract conceptual skeletons identified by Jackendoff and Gruber-concepts like thing, path, goal, and force that maintain connections to physical reasoning but function as abstract symbols. These concepts support inference across domains: if we understand how objects move in space, we can apply similar reasoning to possession, time, or states.
The usefulness of metaphors comes from their analogical structure-they support reasoning of the form "A is to B as X is to Y." This explains why metaphors are "metaphors we live by," enabling sophisticated inferences like the engineers who solved synthetic paintbrush problems by reconceiving brushes as pumps.
Capítulo 7
The Dance of Indirection
How do we read between the lines in conversation? Philosopher Paul Grice explained this through his theory of conversational implicature. When a reviewer writes that a performer "produced a series of sounds that corresponded closely with the score," we understand it as criticism because the verbose phrasing deliberately avoids the simpler word "sang." Grice proposed that conversation follows a "cooperative principle" governed by four maxims: Quantity (say just enough), Quality (be truthful), Manner (be clear and brief), and Relevance.
Politeness theory explains how people navigate social interactions through language. Anthropologists Brown and Levinson showed that people worldwide worry about maintaining "face"-both positive face (the desire to be approved) and negative face (the desire for autonomy). These reflect fundamental social dualities: solidarity/status, connection/autonomy, intimacy/power.
Most speech threatens the hearer's face somehow. Simply starting a conversation demands attention; commands challenge autonomy; requests risk refusal. We compensate with politeness strategies: sympathy (simulating closeness through phatic expressions like "How are you?") and deference (acknowledging the hearer's autonomy).
Game theory helps explain why people use indirect speech in situations of conflict or uncertainty. Consider a driver deciding whether to bribe an officer after being pulled over. A direct bribe ("I'll pay you fifty dollars to let me go") creates a risky binary outcome: if the officer is corrupt, the driver avoids a ticket; if honest, the driver gets arrested for bribery.
Using an indirect bribe ("Maybe the best thing would be to take care of it here") creates a safer third option. A corrupt officer understands the implied bribe, while an honest officer can't prove bribery beyond reasonable doubt. The indirect approach combines the advantage of bribing a dishonest cop with minimal risk when facing an honest one.
Why does bribing a maitre d' feel unethical when we readily pay for expedited service elsewhere? Anthropologist Alan Fiske explains this through his theory of human relationships, which identifies three fundamental types.
The first is Communal Sharing, where "what's mine is thine." This relationship naturally exists among blood relatives due to evolutionary advantages, but also extends to couples, friends with shared interests, and others with common goals. Communality isn't calculated through genetic relatedness but experienced as emotional warmth and conceptualized through metaphors: SOLIDARITY IS PHYSICAL CLOSENESS, SOLIDARITY IS BEING CONNECTED, and most powerfully, SOLIDARITY IS BEING MADE OF THE SAME FLESH.
Since pure communality is vulnerable to exploitation, groups employ "mind-control techniques" to strengthen communal bonds. These include kinship metaphors (brotherhood, sisterhood), shared meals and food taboos, blood-mingling rituals, distinctive body modifications, and synchronized movements that create the sensation of being part of a single communal body.
Capítulo 8
The Paradox of Transparent Pretense
Why do we bother with transparent pretenses in social interactions? When the maitre d' knows he's being bribed, the donor knows he's being solicited, and everyone understands a sexual overture, why does keeping these intentions "off the record" matter if the deniability wouldn't pass "the giggle test"?
This puzzle emerges when two conditions are met: both parties know each other's intentions, and the implicature is so obvious that there's no real doubt about what was intended. In a courtroom, even a scintilla of deniability might matter legally, but why do we act like defense lawyers in everyday life?
The phenomenon reveals something specific to language itself-expressing a sentiment baldly, on record, in so many words makes a psychological difference. "Some things once said can never be unsaid."
Several explanations might account for this peculiar human tendency:
1. The Token Bow: By using indirect speech, a speaker signals they're making an effort to spare the hearer's dignity or face. This consideration itself eases feelings all around, while a bald proposition signals no such effort.
2. Don't Talk At All; Show Me: Communal relationships aren't negotiated verbally but consecrated through physical signs like rituals and bodily contact. The very act of articulating a relationship in words signals it can't be communal, since true communal bonds are felt, not rationally decided.
3. The Virtual Audience: While participants may understand an indirect proposition perfectly, eavesdroppers or third parties hearing about it secondhand have only the actual words to go on. We may always be performing for this imagined audience.
4. Preserving the Spell: The public relationship is a pleasant illusion-the maitre d' reigning over his fiefdom, a woman enjoying a man's intellectual interest, a donor enjoying conviviality with prestigious people. This illusion can survive an implicated message but would be shattered by a bald proposition.
5. Certainty as a Focal Point: Relationship types represent discrete, very different modes of interaction, and switching between them requires joint recognition. A bald proposition clearly falls on one side of the line, while innuendo exists on a continuum. Even if the plausibility of denial is tiny, as long as it isn't absolutely zero (as with a bald proposition), it provides a clear boundary for maintaining the current relationship.
Language is fundamentally social-we choose words carefully to both convey intentions and maintain relationships. But is our avoidance of plain speaking a design flaw or might it serve a deeper purpose?
The paradox of rational ignorance suggests that even if our brains could process unlimited information perfectly, there are certain messages a rational mind might deliberately avoid receiving. Knowledge isn't always power-sometimes it's poison.
Capítulo 9
The Kantian Framework of Human Thought
Our minds impose a causal texture on the world that goes beyond mere temporal sequence. When my PDA chimes a minute before my alarm clock blares each morning, I never perceive the chime as causing the alarm despite their reliable sequence. Instead, I attribute the alarm to my button-pressing the night before-a cause separated by hours, with an uncertain mechanism, and occasional failures.
Space, time, and causality form the fundamental infrastructure of our experience, yet they're riddled with paradoxes. I cannot imagine an elephant without spatial properties-it must occupy space somewhere, even if floating in my mind's eye. Similarly, we can imagine empty time but not events outside of time. Space itself presents a logical paradox: it must be either finite or infinite, yet neither possibility sits well with intuition.
These paradoxes led Kant to conclude that space, time and causality aren't properties of reality itself but structures imposed by our minds. They organize our experience rather than being abstracted from it. We grasp the world only through these mental frameworks, which harmonize with reality well enough for science to be possible, but prevent us from knowing "the world in itself."
Kant's conceptual scaffolding that organizes our experience is conspicuously reflected in language structure. Rather than being organized around sensory experiences or human obsessions like food or sex, languages are structured around abstract Kantian categories: substance in nouns, space in prepositions, causality and time in verbs. We see this in everyday metaphors too-prices "rising" like balloons, cities being "an hour apart" as if they were alarm clocks.
This isn't to say Kant is a reliable guide to our current understanding of thought and reality. Modern physics contradicts our intuitive experience of space and time, showing space to be warped, possibly eleven-dimensional, and riddled with black holes. Time may be just the fourth dimension of a static space-time. Many physicists say space and time as empty media don't exist at all.
Yet two Kantian ideas remain invaluable: his synthesis of empiricism and rationalism (the mind contributes abstract conceptual frameworks that organize experience but must be filled with sensory content), and his characterization of space and time as media in which sensations are arrayed.
Capítulo 10
The Human Mind's Conceptual Toolkit
Humans transform the analog flow of sensation into discrete objects and events, packaging experience into propositions about real and possible worlds. These characterizations are schematic, highlighting certain aspects while ignoring others, allowing multiple construals of the same physical situation.
Our mental inventory begins with basic units-events, states, things, substances, places, goals-and specifies how these units interact through actions like causing, enabling, and preventing. We distinguish individuals from categories, essences from superficial properties, and substances from bounded objects. Our primitive number concept distinguishes only one, two, and many, yet applies across quantities, space, and time.
Nouns reflect profound distinctions in how we mentally represent entities, particularly through the count-mass distinction. Count nouns like "pebble" refer to bounded objects that can be enumerated, while mass nouns like "gravel" refer to substances without clear boundaries. The grammar enforces these distinctions: we say "two pebbles" but not "two gravels," "a pebble" but not "a gravel," and "much gravel" but not "much pebble."
Spatial terms quantize space at points where causal events change dramatically on either side. The point at which we say a marble is "in" rather than "on" a hand corresponds to the curvature that prevents it from rolling off when jiggled. A rope "around" a pole can do things a rope merely "by" the pole cannot. Such distinctions can be literally life-saving-as when a miscommunication about whether someone fell "on" versus "through" ice fatally delayed rescue workers.
Time is the essential fabric of consciousness-while we might imagine existing without space (as if in sensory deprivation), we cannot conceive of awareness without time. Despite claims to the contrary, all human cultures have conceptions of time; no society exists where people would be unsurprised by someone aging backward.
Causality, as Hume noted, is "the cement of the universe." Our intuitive understanding of causation goes beyond simple correlation or counterfactual theories, involving a sense of "oomph"-some force or energy transmitted from cause to effect.
Kant was right that our minds "cleave the air" with concepts of substance, space, time, and causality. These concepts form the foundation of conscious experience and provide the semantic content for major syntactic elements: nouns, prepositions, tense, and verbs. They give us the mental vocabulary to reason about physical and social worlds.
However, when examined through language, these concepts differ significantly from what Kant imagined. Rather than matching reality perfectly, they are "gadgets in the brain" that present paradoxes when pushed to the frontiers of science, philosophy, and law. They are digital where the world is analog, schematic where reality is rich, vague where precision is needed, and parochial to human interests when objectivity is required.
Though it may be humbling to see the foundations of common sense as merely design specifications of an organ, our science and reason have managed to reveal aspects of substance, space, time, and causality that transcend common sense. This scrutiny of concepts as they appear in language allows us to understand them as part of our makeup and discount them accordingly-coming as close as possible to the light dove's dream of taking wing in empty space.
