Le mieux est l’ennemi du bien.
Voltaire
Not everything worth doing is worth doing well.
Kenneth L. Pike
NOBEL-PRIZE-WINNING economist Herbert Simon introduced the concept of ‘satisficing’ into the science of problem solving. His point was that the solutions preferred by business, throughout human endeavours and by the mind itself are not usually the best ones but the good-enough ones – the ones that ‘satisfice’ the need rather than perfectly satisfy it. The same principle applies to evolution. With respect to language, this means that human grammars and sound systems are not required to be optimal – in fact, they never are. Language gets the job done just well enough, never perfectly. Herbert Simon has echoed Voltaire, claiming: ‘The good is the eternal enemy of the best.’
It is this aspect of language that so strongly supports the idea that it is an ancient invention, tinkered with throughout human history. There are basic strategies of communication that usually work, but frequently fail – such as when one omits information that is assumed to be shared by one’s interlocutor. Or communication can fail due to the failure to remember or the very lack of a word or, more importantly, a sentence, to translate a concept in one culture or person or language into one’s native language. Human languages leak. They are not mathematical, perfectly logical codes.
If someone yells, ‘Stop the car at the stop sign!’ they assume that the person they are instructing knows what ‘stop’ means, how to drive a ‘car’, what a stop sign is and what it means to stop at a stop sign as opposed to stopping on the edge of a cliff. (One may stop at a sign with the front wheels slightly beyond the sign. That wouldn’t work out well at a cliff.) Those assumptions are just built into the choice of the words. I imagine that most readers of this book, as members of a largely uniform driving culture, know that one has not ‘stopped at the stop sign’ if they stop 200 yards in front of it or ten feet in front of it or even if it is level with the rear seats. Proper stoppage at the signage brings the vehicle to a complete stop roughly 1–5 feet before the front of the car passes the stop sign. This is part cultural knowledge, part lexical (word) knowledge. ‘Common’ sense is just experience and acquired cultural information.
For a more involved example, consider a 2016 opinion piece published in the New York Times on the purported hacking of a presidential candidate’s emails by Russia. Most of the values embedded in this opinion piece are easy to discover, though some are worth pointing out for discussion because one often reads over such unspoken information without noticing it. A few comments are offered, in square brackets in order to reveal unspoken cultural and contextual information and values that would be expected to be implicitly understood by every culturally North American reader. Imagine the potential effect of seeing such editorials in a newspaper one takes on a regular basis over time, reading them passively, perhaps, while sipping coffee at the breakfast table or on the train or bus to work. Their effect could be largely subliminal, reinforced with every story that assumes similar values, which is the case of particular newspapers, with the New York Times usually speaking for liberalism.
The Real Plot Against America
Timothy Egan – July 29, 2016
In retrospect [By beginning with a reference looking back, the author appeals to the reader’s assumed knowledge of what is to come], it [Because the pronoun ‘it’ is used, the author tells the reader: ‘I assume you know this’ – that which will become clear] worked out much better than planned. [It still is not explicit what the author is talking about. This assumption that the reader shares the author’s knowledge builds a potential bond – ‘we’re in this together!’] Who’d have thought a pariah nation, [the author assumes that the reader knows what a nation is. And the use of the world ‘pariah’ is a value judgment that may not have been shared by all readers, but since the author is assuming a bond between himself and the reader, the reader will likely agree with this judgment about a sovereign state that in fact has relationships with most nations of the world] run by an authoritarian [This is a reference to the president of Russia, Vladimir Putin, who enjoys an 80% approval rating among the Russian population – ‘authoritarian’ is a value judgement that is not shared by all] who makes his political opponents disappear, [Putin is a murderer] could so easily hijack a great democracy? It didn’t take much. A talented nerd can bring down a minnow of a nation. But this level of political crime requires more refined mechanics – you need everyone to play their assigned roles. You start with a stooge, a fugitive holed up in London, releasing stolen emails on the eve of the Democratic National Convention, in the name of ‘transparency’. Cyberburglars rely on a partner in crime to pick up stolen goods. And WikiLeaks has always been there for Russia, a nation with no transparency. [This entire passage is full of value judgements that are not universally shared but which the author believes are shared by the readers of the New York Times] …
The point here is not that newspaper pieces have deep and profound shared cultural knowledge (though what is ‘deep’ and ‘profound’ can vary from culture to culture and reader to reader). The point, rather, is simply that the author and the reader, for any of this opinion piece to work, must share cultural knowledge and, preferably, also a similar set of values. Or, if they do not share the values, both are adept at interpreting the values ‘hidden’ among the words.
Any article or opinion piece is, like this piece, saturated with unspoken judgements, opinions, values and knowledge that are never stated. The ‘underdeterminacy’ of this kind of information, its implicit nature, brings us full circle to the Banawá conversation this book began with. In human interaction the unstated is always crucial. Without culture, there is no language.
The British philosopher Paul Grice developed some helpful concepts for understanding the cultural and communicational presuppositions that underlie all human communication, which he referred to in the aggregate as the ‘cooperative principle’. As Grice said, summing up his ideas, ‘Make your contribution such as it is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged.’ Grice makes this look like advice or a command, but it is actually intended as just a description of the cultural conventions underlying communication. We don’t have to be taught these things. This is how we behave.
More accurately, Grice’s principle of cooperation in communication is how we operate if we actually want the person(s) to whom we are talking to understand us. Every adept speaker follows the cooperative principle, as does every adept hearer. Their assumptions are further built on their unspoken cultural knowledge. Grice divided his cooperative principle into several ‘maxims’ which, when observed and, perhaps especially, when flouted allow for what philosophers and linguists refer to as ‘conversational implicatures’, things unsaid that are crucial for the meaning of what we have heard or spoken. The four maxims of Grice’s principle caught on quickly among linguists, philosophers, psychologists and social scientists. These maxims are: the maxim of quality, the maxim of quantity, the maxim of relevance and the maxim of manner. And they are the perfect kind of discovery – simple and intuitively right.
The maxim of quality assumes that everyone will speak the truth. It presumes that neither the hearer nor the speaker will believe that anything will be presented as true if it is known to be false. It also assumes that no one will say that something is true if they lack adequate evidence. There are, of course, lots of untrue things said and many spurious postings of made-up facts on the internet. So Grice is not saying that people cannot lie. He is saying that hearers assume that they are not being lied to in the normal course of interaction.
In fact, in many languages, such as Pirahã, the actual verb in the sentence has to have a suffix that tells the hearer how good the evidence is for what the speaker is saying – inference, hearsay, or direct observation. So if one Pirahã asks another, ‘Did so-and-so leave the village?’ one possible response is, ‘Yes, he did,’ where the verb did would have a suffix appended that might indicate ‘I saw him leave’ or ‘Someone told me he left’ or ‘His canoe is not here, so I infer that he has left’ and so on.
To lie in any language, therefore, is to disregard the maxim of quality. Of course, since everyone lies, we know that there are times when we intentionally flout the cooperative principle. But even though we know that others flout the maxim and even though we know that we ourselves also flout it, if you tell someone something, they will initially believe it, other things being equal. In fact, English, like Pirahã, has verb forms that indicate the degree of truth or certainty in the things that we say. In English we call these markers ‘moods’. So there is the indicative mood: ‘John went to town’. There is the subjunctive mood: ‘If John were to go to town’. There is the conditional mood: ‘I wish that you would leave’. Or the imperative mood: ‘John! DO IT!’ All of these in their own way express the relationship of the word meanings to the truth of these meanings applied to the world around them. So indicative means that the world is the way it is being described. The subjunctive means that the speaker imagines that the world could possibly at sometime be the way it is being described. The conditional mood means that one would like to see the world in a particular way or does not want to see the world in particular way. The imperative means that one wants the hearer to make the world a certain way that it currently is not.
The next of the four maxims Grice lays out is the maxim of quantity. This is again in two parts. First, don’t give any more information than the exchange requires. Second, relay all the information necessary for the current interaction. Let’s say that someone asks, passing in the hall, ‘Hey, how ya doin?’ And you answer, ‘At 8.30 I have a dental appointment. I have an irritable bowel issue today. I spent last night worried about my finances. Other than that, OK.’ Or someone says, ‘How did you meet your spouse?’ And you answer with every detail you can remember about the concert at which you met. In both cases, English has a phrase to describe your answer: TMI – too much information. These answers exceed the information requested! This occurs when a speaker confuses irrelevant with relevant information. TMI violates the maxim of quantity. But there is another way to violate the maxim – too little information. Imagine that someone asks, ‘What do you want to do tonight?’ And imagine that the response is, ‘Whatever.’ Well, although that is not an unheard of answer to such a question, it is unhelpful. Such a vague answer fails to provide the amount of information expected in the exchange. Giving too much information or too little information deliberately are examples of flouting the maxim of relation (or relevance).
One of the more famous examples of defying a maxim is the letter of recommendation. Someone writes one but manages to say very little about the candidate’s qualifications. They might offer the judgement that ‘John has excellent handwriting’. Everyone knows in this case that the writer is flouting the maxim of quantity and implying that John is not qualified. What happened here? How does this implication emerge from the literal meaning of the words in the answer?
Or consider a spouse’s violation of the maxim of relevance:
Husband: ‘How much longer will you be?’
Wife: ‘Mix yourself a drink.’
To interpret his wife’s answer, the husband assumes first that she is following the maxim of relevance. Her answer, though it seems irrelevant, must be relevant. She is flouting the maxim (and maybe her husband’s expectations as well). To understand how this otherwise irrelevant comment could be relevant, the husband must go through a set of cultural and personally based inferences. The husband concludes that his wife heard him and understood his question and that her answer, while not literally a response to the question, indicates that he should relax, she’ll be ready in plenty of time. He shouldn’t worry or bother her further. And the wife, for her part, has to know that the husband will be able to draw these inferences, based on her own inferences of how he will interpret her. Both of these examples – handwriting and fix a drink – work because they flout the maxim of relevance, ‘be relevant’. Implicatures, how people interpret the flouting of maxims, are cognitively complex. They draw on a store of background cultural knowledge. For this reason interpreting conversation in light of the cooperative principle is highly culture specific. The maxims themselves, on the other hand, are probably found in all languages. Grice’s maxims do not supplant culture. They assume it.
Consider Grice’s maxim of manner. The interlocutors assume that each intends to be clear in their speech. ‘Being clear’ in this sense has four subcomponents. First, avoid obscurity. People believe that a speaker is making an effort to avoid ambiguity, to be as brief as possible while respecting the maxim of quality and to be orderly in their remarks. Again, these are not rules of etiquette for speech. Grice’s claim is that his maxims are assumed by everyone when they talk. If someone uses an obscure expression, therefore, when their hearer expected a clear expression, they must mean something non-literal – they must be flouting this maxim for a purpose. So one infers the speaker’s meaning. And if they come from the same culture or know each other well, they will in all likelihood infer correctly. Not always, however. There are frequently bad inferences that lead to confusion or misunderstanding.
People also interpret others charitably or uncharitably. That is, we believe someone means something good when we interpret them charitably. This is a favourable bias towards them or the situation’s likely meanings. If someone says, ‘That’s an ambitious statement,’ and their hearer interprets them charitably the hearer will assume that what is meant is something like, ‘You really know your stuff. You are going places!’ But if one interprets the same remark uncharitably they will quite possibly take this to mean, ‘You bit off more than you could chew and your statement failed.’ People use these modes of interpretation frequently in politics. They tend to interpret their own candidate charitably and their opponent’s uncharitably. These types of interpretation are found all around – in marriage, sibling relationships, work and so on. The way people interpret what someone is saying is based to a large degree on the kind of relationship they have. A standard joke heard among university administrators is, ‘Gee, if I say “good morning” to so-and-so, they will ask themselves “I wonder what he meant by that?”’ If an employee either does not trust or fears his supervisor, this will colour his interpretation of what the supervisor says, however innocuous the supervisor’s intent. If one believes in someone, trusts them and values their friendship, then if that person says, ‘I will find you, no matter where you are,’ the hearer will at least believe that the speaker will try hard to find them. If someone says, ‘When I am elected president, I will make America more secure,’ one is less likely to believe them. This is in part because they aren’t known personally, or because no one believes any politician about anything. At least, they will be less credible than a ‘normal person’ would be, no matter what they are talking about.
Likewise, one’s cultural experiences (however valid intellectually) can affect their interpretation of groups as well as individuals. If someone believes that all rich people are corrupt, then they will be less likely to believe someone rich when they say that the ability to make lots of money is good for the entire community, even if only one person does make a lot of money. If one believes that anyone who receives government assistance, welfare of any kind, is lazy or irresponsible, then if such a person says, ‘I have to lie down,’ one may be more inclined to see that as laziness than illness or being legitimately tired from hard work, even if one otherwise knows nothing about the person speaking.
This is all crucially relevant to language evolution. Even if erectus only said things like ‘Eat. Drink. Man. Woman?’ another erectus would have had to know what woman or group of women the speaker had in mind, when he might want to eat, whether he was telling him to get out of the way of his plans, among much other assumed information. Language is underspecified for meaning. Without culture, whether for sapiens or erectus, there is no communication. When he proposed the cooperative principle Grice revealed something about language evolution therefore that even he was in all probability unaware of. Only creatures that follow it can have language. There is no need to defend or criticise how we interpret others by a principle of charity. It is just a crucial characteristic of psychology in many cultures.
The relevance of all the above to language evolution is that even Homo erectus, Homo neanderthalensis, Denisovans and Homo sapiens would have – in the gradual construction of relationships, roles and shared knowledge bases – interpreted what people said, from the very first syllable uttered or gesture made, based on their view of the person and their understanding of their context. They would have ‘filled in the blanks’ of speech just as sapiens do. This is all a part of language that many linguists call pragmatics – the cultural constraints on how language is used. And these constraints guide our interpretations of others. They help us, as they helped other Homo species, to resolve the underdeterminacy of speech.
Another example of language being just good enough, depending on cultural knowledge for its use, is found in ‘speech acts’, the use of language to accomplish specific kinds of cultural goals. Oxford don John Austin and his student, Berkeley philosophy professor John Searle, introduced the analysis and terminology of speech acts into discourse about human language. Whenever anyone speaks to someone else they are engaged in an action of a very particular type. In fact, they are simultaneously occupied in many distinct acts. Austin talked about locutionary acts (what was said), illocutionary acts (what was meant) and perlocutionary acts (what happened as a result of what was said and what was meant). Each of these is important to the understanding of the nature and use of language and is therefore important to the understanding of the origins of language. And each of these must have been a feature of language since its beginning.
The locutionary act is speaking itself. If one asks, ‘Where is Bill?’ the very moving of the mouth, emission of air from the lungs and the arrangement and selection of the words used are all part of the locutionary act. But anyone performing this locutionary act is simultaneously performing an illocutionary act. An illocutionary act is the effect one intends their utterance to have. If one promises something they want their hearer to recognise that their promise is a promise. That is the effect one intends for their words to have. The illocutionary acts a person’s words can accomplish include statements, commands, questions, or performative acts. The latter occurs when a minister, legally authorised to perform marriages, concludes a legitimate (non-faked, non-Hollywood) marriage ceremony with the words ‘I now pronounce you man/husband/wife and wife/husband’. Many cultures make it such that uttering those words in the right context with the right authority (what philosophers refer to as the ‘conditions of satisfaction’ of the act) legitimates the couple’s marriage. That is a performative act. These acts require more specific cultural supports than either statements or questions do. Therefore, they in all likelihood would have appeared later in the evolutionary record.
As languages have evolved, they have come to possess various types of illocutionary acts. One type is called representatives – acts that commit the speaker to the truth of the content of what she is reading or saying, such as a witness taking an oath in court or all graduating seniors reading the college pledge together. Another is referred to as directives – acts to get the hearer to do something. Directives include exhortations, direct orders, advice and requests. Then there are commissives, which are acts of commitment by the speaker, including promises and oaths of office. Expressives communicate attitudes and emotions, such as congratulations or apologies. Performatives are acts that by their mere performance bring something about, such as a judge passing sentence. The list of speech acts recognised by researchers varies depending upon the author. But what is important about them for language evolution is that they show how the use of language is anchored in culture. Homo erectus is likely to have used representatives, directives, questions and commissives.
Finally, there is the perlocutionary act – what happens, or what one hopes will happen, in the mind of my hearer when they speak. At the end of an attempt to persuade someone, the effect of that person actually being persuaded or unpersuaded is a perlocutionary act. Thus a translator of a book might say that a good translation should produce the same perlocutionary effect in the readers of the translation that was produced by the original. In other words, we communicate for effects, perlocutionary acts. If one speaks or translates or otherwise engages in the communicative enterprise they are hoping that their locutionary act will be the right choice for the right illocutionary act to produce the desired perlocutionary effect or act.
Notice that there is no way to have a language without perlocutionary, illocutionary and locutionary acts. If Homo erectus talked, then they performed these acts. They would also eventually have learned to be polite.
A neanderthalensis could have just grunted and demanded a piece of meat. And that might even be what most of them did, because they might not have had request forms as well developed as our own, at least not initially. Politeness is interpreted as indirectness and gentleness in the manner of letting people know what it is that one wants them to do. It can also be used to report on the condition of one’s body (such as, ‘It’s been a long time since we’ve eaten,’ or, ‘Where’s the bathroom?’ both of which indirectly report on a need), and seems to be sufficiently subtle and nuanced that it would have come much later in the development of language. As people began to learn that the use of force was usually inefficient when interacting with those in their group they would have begun to rely instead on the use of persuasion to get what they wanted. The rise of politeness, adroitness and foresight in getting the other to want to help us, or at least not feel forced to help us, led to the evolution of indirect speech acts. These can take the form of speech or gesture or body language, but their function is the same – getting another to do something without actually saying what one wants them to do. This is another example of the central truth about language, already stated earlier, namely that we do not say what we mean and we often do not mean what we say.
For instance, a man might prefer the air-conditioning thermostat to be set just about cold enough to hang raw meat, but his wife likes the temperature considerably warmer. She, being many times more polite than he is, let’s suppose, might ask, ‘Aren’t you cold?’ Or, ‘What temperature do you have the thermostat set at?’ Or even, ‘I am cold.’
The man is supposed to infer, obviously, from the mere raising of the topic of temperature that someone, his wife, is dissatisfied and that he is being asked to do something about it. There may be body language accompanying such indirect requests that make the purpose behind the words even clearer. She may wrap herself in a blanket ostentatiously. But the main principle is what has been seen over and over – people never say all they mean. Hearers must use both knowledge about language and cultural knowledge (about thermostats in this case) to respond. If the man ask his wife, ‘Do you want me to turn the thermostat up?’ she might very well reply, ‘No, I am OK.’ But woe unto him if he does not figure out quickly that what she actually means is, ‘Of course. Do I have to draw you a picture? Get a move on.’ Of course, her indirect way of getting her request across is much more effective than a direct command.
I overheard a well-known philosopher and logician at a major university explain that if one was really angry with someone they could punch their lights out without fear of legal consequences. He explained further, only half-humorously, that if there are witnesses and one wants to get away with giving someone a drubbing, they should say exactly the opposite of what they’re doing. Push someone over and say, ‘Oh, you fell! Let me help you up.’ Then one might kick them in the head while saying, ‘Oh no, I tripped.’ Then one might give them an elbow to the nose while uttering, ‘Watch out! You’re falling again. I have you! Don’t worry!’ The courts might see through this, but there is a chance that when the witnesses report what they saw the defendant’s attorney will ask them what they heard. And then the prosecution case might fall apart. Why? The answer is not hard. Court systems often assume an inadequate theory of language, a theory that ignores what is not said and focuses only on what is actually uttered. But often what is not said is still communicated forcefully and what is said is just a decoy for something else.
The lesson, perhaps with some overkill, is again that language engages the whole person and the whole culture. In fact, it is more serious than this. One can make a case that no one can fully understand what anyone else says. We understand just well enough to get by. Or, as Herbert Simon said, language is just good enough – it ‘satisfices’ our requirements, but it is by no means the perfect system of communication. Yet when working hand in glove with culture, language is incredibly complex and rich. And such complexity and depth could only result from the evolution of the body and brain in tandem with psychology, language and culture for hundreds of thousands of years.
Conversational implicatures do not exhaust the contributions of the context to the interpretation of utterances, implicit information and speech acts, however. But they allow grammar to specify less information than required, leaving it to speakers to infer the remainder of the meanings from the context and culture of their exchanges. Grammar and culture work together in modern languages and this conjunction was almost certainly vital for the development of earlier human languages into modern languages. The grammar helps the inferences and interpretations of speakers, it does not determine them.
As linguists, philosophers and psychologists thought more about Grice’s work on the cooperative principle, they discovered additional ways in which cooperation helps to structure language. In the mid-1990s Dan Sperber, a cognitive scientist at the CNRS (Centre national de recherche scientifique) in France, and Deirdre Wilson, Professor of Linguistics at University College London, collaborated to produce the theory of human interaction known as ‘relevance theory’. Relevance theory explores applications of the cooperative principle beyond conversation. Relevance theory, like Grice’s work, sees language forms and interactions as governed by a culturally based pragmatics.
All of the work in pragmatics (which is partially the study of how the context in which one is conversing determines which interpretations are appropriate), sociolinguistics (how language and society each affect the other) and additional disciplines of scientific research that look at language in use, rejects the so-called ‘conduit metaphor’ of language. This represented a tremendous advance in the formalisation of the theory of communication. And the highly respected name usually associated with its formulation is mathematician Claude Shannon. This metaphor is the idea that communication is linear, where a thought arises in the speaker’s mind. Next, the speaker selects a grammatical form for that idea. The speaker then transmits the idea-in-the-form to the hearer’s ears (or eyes, if using sign language). The hearer takes apart the grammatical form and is left with the meaning that the speaker intended. This is represented in Figure 35.
Figure 35: Shannon’s conduit metaphor of communication
In the late forties and early fifties, when Shannon was writing his most pioneering work, there was almost no research on the formalisation of communication theory (though certainly work such as that of Alan Turing on the code of Nazi Germany’s Enigma machine was foundational to thinking in the area). As a researcher at Bell Labs it was Shannon’s job to work out an understanding of communication that could be translated into mathematical models in order to help Bell make telephones ever more effective and efficient.
As Figure 35 shows, Shannon’s conceptualisation of the communication problem leaves no space for external influences on the process of communication, such as culture or context or gestures or intonation. It is almost as though all one needed to communicate was two brains, two vocal apparatuses and two auditory systems. Shannon, who knew Alan Turing and other founders of computational theory, developed his system in 1948 and it has been accepted as foundational to most other work in the fields of cognitive science, electrical engineering, linguistics, psychology, mathematics and other fields since then.
But Sperber and Wilson, following on the work of Grice, Searle, Austin and many others, said in effect, ‘The conduit metaphor has had its day, but it really doesn’t capture what happens in human communication at all. The conduit is just a set of points in a much larger set of events and processes that underlie communication.’ In the view of relevance theory, any time a story is told or a conversation is had or a sentence is uttered, there is always a context of utterance. Not only this, the set of interlocutors assumes that each member of a particular communication event – a storytelling, a speech, a conversation – knows the context as well as the relevance of the context to what is being said and is therefore able to comprehend and respond appropriately, in a relevant manner. One doesn’t say something unless it is relevant. One doesn’t imply or interpret something unless it is relevant to the context in which the discussion is taking place. Thus if someone says something to someone else, the hearer will assume that it is relevant and they will therefore make an effort to understand what was said or written.
As a concrete example, what is the relevance of a discussion of relevance theory or Grice’s cooperative principle in a book about language evolution? The reader makes the effort to process the sentences that they are reading but does so assuming that these sentences will be connected in some way to our general topic of discussion – language evolution. And indeed they are. But before turning to the evolutionary significance of these ideas, one more item on the topic of ‘language emerges from context’ should be discussed – conversation. The apex of our evolution as a species.
Conversation seems innocuous enough. I say something. You say something. We think about what the other has said, or not. Then we finish and say goodbye. Something like that. This simplistic view of conversation is not entirely wrong. It is just fundamentally incomplete. Therefore, to understand how language evolved to be ‘just good enough’ we need to look at conversation. Here is a snippet of conversation from a couple I observe regularly:
WIFE: What time are we going to pick up Miguel on Thursday?
WIFE: Yes, but that is when you were planning to pick him up after work. Now we are just driving directly to pick him up.
HUSBAND: The time doesn’t change as a result.
WIFE: But just give me a time that we will need to leave, so I can call the dogsitter.
HUSBAND: Whenever you want to leave in order to get to his place between ten and twelve.
WIFE: Why can’t you just tell me the time? Why are you playing these games?
HUSBAND: I gave you a range and from that there should be no problem inferring the time.
WIFE: I don’t even care if we go.
HUSBAND: Fine.
The interpretation of this depends on the literal meanings of the words, but also on the personalities involved, the cultural notions of time, the complication of the drive to pick up the person in question and the fact that both people are tired. One wants things stated clearly and precisely. The other likes to be less committed to times (in some situations – in other situations their roles are in all likelihood reversed with regard to time and precision). Moreover, when the wife says, ‘I don’t even care if we go,’ that is not intended literally. Both interlocutors are keen to spend time at the beach with the friend being discussed. This line is uttered to make a point, namely, ‘If you cannot simply answer my question with a specific time, then my feelings must not be that important to you.’ The final, petulant response, ‘Fine,’ is also not literal in this case. It is just to say, ‘If you are going to make a big deal of this, I am not going to respond as expected.’ Aside from their amusement value, such exchanges show how the interpretation, the utterances and the exchange as a whole can only be understood through extensive external knowledge of culture, local circumstances, the relationship between the interlocutors and their individual personalities. Shannon’s conduit metaphor is of very little help here.
Language, psychology and culture, again, co-evolved to produce the contextual linkage between the world, personalities, cultural understandings, current events and so on that make full interpretation of language possible. Moreover, we see a great deal of variation in how this is done across different languages. Consider a few exchanges and then a discourse on how to make arrows in Pirahã
Greeting in Pirahã:
Ti soxóá.
‘I already.’
Xigíai. Soxóá.
‘Ok. Already.’
Or this exchange:
Ti gí poogáíhiai baagábogi.
‘I give you a banana.’
Xigíai.
‘OK.’
Or another, this one to communicate that you are leaving, where English speakers might say something like, ‘I am leaving now, goodbye.’
Ti soxóá.
‘I already.’
Gíxai soxóá.
‘You already.’
Soxóá.
‘Already.’
Pirahã doesn’t have words for thanks, goodbye, hello and so on, what linguists refer to as ‘phatic’ language. They allow the context to determine the bulk of the meaning for obvious things like leave-taking and arrivals. They see no need to say thank you, in part because every gift carries an expectation of repayment. If I give you a banana today, you should give me something, such as a piece of fish, when you are able. This is not stated. It is culturally presupposed. So ‘giving’ is actually a form of exchange by and large. If a gift in the English sense is intended, that is, without expectation of repayment, I would say:
Ti gí hoagá poogáíhiaí baagáboí.
I you contraexpectation banana give (literal translation)
‘I (against your expectations) give you a banana.’
The addition of the word hoagá, ‘contraexpectation’, renders the meaning of the entire sentence in effect: ‘Contrary to what we normally do, I am just giving you this.’ This may sound a bit circumlocutious. But it results from the Pirahãs’ assumptions being quite different from those of most Western cultures. In any case, there was no need in the course of the historical development of the Pirahã language to develop vocabulary for such purposes for exchanges, only for giving without expectation of repayment.
When Homo erectus began to communicate orally, it would have been even less necessary to develop a language that expressed everything. After all, language would have emerged from a society that was increasing in complexity and communicating with body language, perhaps vocal interjections, hand signals, maybe some representative spatial drawing with a stick in the dirt as they planned hunts. But let’s assume that a local erectus community somewhere had settled on the rudiments of some symbols. One cannot expect the inventors of a single symbol to try to pack all the meaning in the world into that symbol. In fact, one could not expect that even for a group of symbols, no matter how many symbols in that set. There is too much information in the background context, too much in our memories that we use to interpret but don’t actually say (and often don’t even know we know, or don’t know that we are using) for mere symbols, however enriched by gestures and intonation and body language, to say everything. Thus it is clear that as language evolved, speech acts, indirect speech acts, conversations and stories depended heavily on cooperation, implicit (unspoken) information, culture and context. It is the only way that language has ever worked.