Linguistics is a science and can only be understood as a discipline by treating it as such. The first step, therefore, is to be clear about what science is, and that is what this lecture sets out to explain. The word science comes from Latin scientia, 'knowledge', and in current usage refers to knowledge of the natural world, that is, of the world and the cosmos in which we live. More specifically, science consists of a large collection of best guesses or hypotheses about the natural world which have been tested by observation of the world and, for the time being, are regarded as valid explanations of why the world and the universe are as they are and how they will develop in future. Science makes no claim to absolute truth, however: if further observation contradicts existing hypotheses, these are either appropriately emended or replaced with ones consistent with observation. The discussion is in two main parts: the first part outlines the historical development of science, and the second examines its nature, that is, the philosophy of how science should be pursued. 1. The history of science From the time that historical records begin to appear, and for a long time before that, humans have tried to understand the environment in which they lived partly as a matter of curiosity and partly as a way of controlling that environment. This attempt at understanding has historically taken two main forms: mythology and empirical science. 1. The history of science 1.1 Mythology The earliest known attempts to understand the natural world took the form of mythology, that is, a system of narratives in which aspects of nature were personified as gods and goddesses, and the interaction of natural processes as well as human interactions with these processes were conceptualized as personal interactions among deities and humans. 1. The history of science 1.1 Mythology Religion was and is the ritual whereby humans interact with the gods in the hope of influencing them to provide, for example, good harvests or success in war. Mythology and religion have remained influential in human understanding of the natural world from prehistory to the present day, as the widespread contemporary allegiance to Christianity and Islam as well as to smaller religions attests. 1. The history of science 1.2 Empirical science Empirical knowledge is knowledge gained by observation, and empirical science is therefore knowledge of the natural world gained by observing it. This differs fundamentally from mythology, which explains the natural world by positing a range of unobservable deities and interactions among them and with humans. 1. The history of science 1.2 Empirical science 1.2.1 Empirical science in prehistory As a species, humans are late arrivals in the history of the world. On current estimates, the world is about four and a half billion years old, but creatures that we would recognize as humans rather than some kind of ape have only existed since about 500,000 BC. Our species, homo sapiens, is more recent still, first appearing about 200,000 BC. This means that we have been here for about 0.0001% of the world's existence. 1. The history of science 1.2 Empirical science 1.2.1 Empirical science in prehistory For most of those 200,000 years humans were what archaeologists and anthropologists call hunter-gatherers. They lived in small groups, had a few primitive stone tools, and fed themselves by foraging for edible plants and by hunting animals. To avoid exhausting local resources these groups were frequently on the move; there were no permanent settlements. All of humanity was at the hunter-gatherer stage of development until about 10,000 BC, or about 12,000 years ago. 1. The history of science 1.2 Empirical science 1.2.1 Empirical science in prehistory Apart from what can be inferred from occasional archaeological finds, we have no direct knowledge about the thought-worlds of these early humans, and certainly none about what kind of science they had, if any. Some hunter-gatherer population groups have, however, survived into relatively modern times --for example North American Indians, sub-Saharan Africans, and Australian aboriginals until the 19th century-- and a very few, such as the Amazonian Indians, have managed, just, to survive to the present. From observation of such groups we can infer that the science of prehistoric cultures at the hunter-gatherer stage of development was predominantly or exclusively mythological. 1. The history of science 1.2 Empirical science 1.2.2 Empirical science in early historical times, c.10,000 - c.500 AD The world's temperature is constantly changing, and there have been at least five periods of severe cold when large parts of the earth's surface have been covered by ice. The last of these Ice Ages ended, in the sense that the vast majority of the world's surface became largely ice-free, by about 10,000 BC. 1. The history of science 1.2 Empirical science 1.2.2 Empirical science in early historical times, c.10,000 - c.500 AD At about that time things began to change in one particular part of the world that we now call the Middle East, but was anciently known as Mesopotamia. 1. The history of science 1.2 Empirical science 1.2.2 Empirical science in early historical times, c.10,000 - c.500 AD Here, from about 10,000 BC, wild plants and cereals began to be cultivated and wild animals like sheep and goats to be domesticated. This provided a more predictable food supply, and by about 5000 BC settlements began to appear on the banks of the rivers Tigris and Euphrates 1. The history of science 1.2 Empirical science 1.2.2 Empirical science in early historical times, c.10,000 - c.500 AD The Mesopotamian developments soon spread eastwards into present day Pakistan and northern India and thence into China, and westwards around the Mediterranean basin. Egypt had its pharaohs and pyramids, Greece its acropolis and philosophical schools. About the year 0 the Roman Empire had come into being ,and controlled the territories surrounding the Mediterranean as well as most of present-day Europe. 1. The history of science 1.2 Empirical science 1.2.2 Empirical science in early historical times, c.10,000 - c.500 AD A major development at this time was the invention of writing. The implications of this invention for the development of human culture are huge, and a later lecture examines them in detail. For present purposes, though, its importance lies in the following two observations. • Firstly, it allows human history to be recorded for the first time; in this part of the world at least, we move from prehistoric to historic times. •And, secondly, it made the first developments in empirical science possible. 1. The history of science 1.2 Empirical science 1.2.2 Empirical science in early historical times, c.10,000 - c.500 AD For example: •Traditional knowledge about healing plants and curative procedures was systematized and the earliest form of medicine. •Observation of the heavens for purposes of timekeeping and navigation laid the foundations for later astronomy. •The need for measurement in the construction of cities and various other forms of technology led to the earliest forms of mathematics. The Greeks, in particular, made major advances in geometry. •Greek philosophers from the 7th century BC onwards began to think about how knowledge is acquired, and among other things systematized deductive and inductive forms of reasoning, thereby laying the foundations of modern philosophy. 1. The history of science 1.2 Empirical science 1.2.3 Empirical science in early historical times, c.500 - c.5500 AD By about 500 AD the cultural developments outlined in the preceding section had spread over an extensive part of Eurasia, from present-day Western Europe to China Scientific ideas continued to develop across this area, but to keep the discussion within bounds we will focus on Western Europe in what follows. 1. The history of science 1.2 Empirical science 1.2.3 Empirical science in early historical times, c.500 - c.5500 AD Between about 0 and 500 AD Rome controlled an empire that comprised the Mediterranean Basin, north-western Europe, and parts of the Middle East. This Roman Empire was the culmination of the cultural development described in the foregoing discussion, and prospects were good. Two factors delayed that development for the best part of the next 1500 years, however: 1. The history of science 1.2 Empirical science 1.2.3 Empirical science in early historical times, c.500 - c.5500 AD Over the space of several centuries, groups from northeastern Europe outside the Empire began to attack and then settle within its borders. By 500 AD the Empire in the West had been engulfed and ceased to exist as a political entity. The kingdoms established by these peoples are the foundation on which present-day Western European society is built. 1. The history of science 1.2 Empirical science 1.2.3 Empirical science in early historical times, c.500 - c.5500 AD The problem was that these settlers were at a much lower level of cultural development than that of the Roman Empire. They were unable to understand Roman institutions. As a result these institutions were severely damaged, and it took many centuries for Western European society to recover the Roman level of culture. 1. The history of science 1.2 Empirical science 1.2.3 Empirical science in early historical times, c.500 - c.5500 AD Christianity, originally a small sect of Judaism founded in the decades after the death of Jesus Christ in 33 AD, grew in popularity… to the extent that, by the early fourth century AD, it became the official religion of the Empire. After the end of the Empire it spread into Western Europe and there became ideologically dominant at least until about 1500 AD. This period between about 500 AD and 1500 AD is known as the Middle Ages. 1. The history of science 1.2 Empirical science 1.2.4 Empirical science in Western Europe c.1500 - present The dominance of Christianity in Western Europe included a rigidly enforced ideological orthodoxy: that humankind's place in the natural world was to be understood in terms of Christian theology, that is, mythologically. It was only with the advent of the cultural movement known as the Renaissance between about 1300 and 1500 that this orthodoxy began to be questioned. Intellectuals began once again to base their understanding of nature on observation of it. 1. The history of science 1.2 Empirical science 1.2.4 Empirical science in Western Europe c.1500 - present The medieval view of cosmology, for example, was that the earth was at the centre of the universe and surrounded by a succession of spheres. This was questioned by astronomers like Copernicus (1473–1543), Kepler (1571–1630), and Galileo (1564–1642), who proposed that the earth revolved around the sun. Such developments were fiercely resisted by the Christian ecclesiastical establishment. Despite that opposition empirical science developed rapidly and, in our own time, it offers a highly-developed and ever-developing understanding of the natural world. 1. The history of science 1.2 Empirical science 1.2.4 Empirical science in Western Europe c.1500 - present Mythology remains hugely influential throughout the world, however, and is frequently antagonistic towards empirical science. 2. The nature of science In current Western society, “science” seems to have a special status. If you can convince the right people that the investigation you’re doing is scientific, you get benefits which are not given to people who are not doing ‘scientific’ work. This man just finished a threepart series on his (and others’) work broadcast in prime-time on the BBC (and in high definition). 2. The nature of science Professor Brian Cox says he can explain what’s going to happen in the future, various things that the stars will be doing, etc. The woman who sits in this caravan in the Bigg Market makes similar sorts of claims. However, Angeline Lee is unlikely to be given her own show on the BBC anytime soon (maybe Channel 5 though). 2. The nature of science One of the reasons for this special treatment is that science appears to lead to a special kind of knowledge – true, certain knowledge about the way the world actually is. When scientists make a claim like “sodium chloride dissolves in water” (on the basis of appropriate evidence), it’s not a hope, or a belief. It has some kind of different status. Why is this? What’s special about the investigation that Professor Brian Cox carries out? How is it different from Ms Angeline Lee’s investigation? The philosophy of science tries to get to grips with questions like these, questions like “What is science, really?” and “How is scientific work done?”. 2. The nature of science One of the most important things that the philosophy of science tells us is that many ‘ordinary’ or ‘common sense’ views about science aren’t right. The ‘common sense’ version of science essentially says that it is based on (1) observations (2) of the real world which (3) are built up into generalizations. It turns out that all three of these things require a lot more thought than it might seem at first glance. 2. The nature of science 2.1 There’s no such thing as data: the problem of observation A core component of the ‘common sense’ view of science is that impartial, unbiased, theory-free observation plays a crucial role. Scientists don’t approach their subject with any preconceptions. They simply make observations about what happens in the world and build these observations up into theories that describe the way the world actually is. These observations then constitute the data that the scientific theory tries to explain. 2. The nature of science 2.1 There’s no such thing as data: the problem of observation ‘Pure’ observation However, (visual) observations are not just “there for the taking”. They’re not like camera photos on a retina. Depending on how you look at it, most Westerners will ‘see’ two different staircases. However, the culture of some African tribes does not have the tradition of representing 3D objects two-dimensionally, and they report only seeing a bunch of lines. Whose ‘observation’ is right if observation has a cultural dimension? What’s the ‘data’ here? 2. The nature of science 2.1 There’s no such thing as data: the problem of observation Observation can also be dependent on beliefs and expectations. In a famous experiment, Bruner and Postman (1949) asked subjects to identify a series of playing cards. However, some cards were made anomalous For the normal cards, the identification was almost always correct, but the anomalous cards were, without hesitation or puzzlement, consistently identified as a ‘normal’ card. What’s going on here? 2. The nature of science 2.1 There’s no such thing as data: the problem of observation Observation also appears to be biologically conditioned. A dot is quickly flashed in one location and then another. If you get the timing right, subjects will report that they saw one dot which moved from one position to the other. Thus, it seems like the notion of ‘pure’ observation, the main way data is supposedly gathered under the common-sense view of science, is a problem. At the very least, the question is a lot more subtle than it might seem at first glance. 2. The nature of science 2.1 There’s no such thing as data: the problem of observation Observation is theory-dependent Another important point to acknowledge is that there is no such thing as theory-independent observation, and therefore no such thing as ‘pure’ data. All observations, even the most basic ones, presuppose a theory of the world and the object under observation. 2. The nature of science 2.1 There’s no such thing as data: the problem of observation First, what you choose to observe depends on your theory. In conducting a scientific experiment to measure radio waves, a physicist won’t make a note of the colour of his shirt or the time of day. Why not? If science is engaged in ‘pure’ observation, free of any theoretical bias, then this decision isn’t rational. Why is this factor ignored as opposed to any other? The reason is that the scientist’s theory of radio waves says that those factors aren’t relevant – that they won’t affect the experiment. 2. The nature of science 2.1 There’s no such thing as data: the problem of observation And if these theories are wrong, then the observations need to be discounted. In the late 1800’s, the German physicist Heinrich Hertz did some experiments to measure the speed of radio waves and ran into a problem. Electromagnetic theory said that the waves should travel at the speed of light, but when he and others measured it, they got a different value. It turned out that one thing Hertz wasn’t measuring was the size of the room, because, like the colour of his shirt, his theory said that it wasn’t relevant. However, what happened was the radio waves were bouncing off the walls in the room and hitting the detector, so it turned out that an error in the (unexpressed, implicit) background theory invalidated the observations. 2. The nature of science 2.1 There’s no such thing as data: the problem of observation A similar issue arose in early astronomy. The Copernican view of the solar system predicts that the size of Venus should change appreciably during the course of the year, but observers on Earth didn’t notice any change. However the observation “Venus does not change size during the year” presupposes a theory which says that the size of small light sources can be accurately gauged by the naked eye, and this isn’t true. Modern optical theory explains why this is true and why observations with a telescope will be more appropriate. (And these show that Venus does indeed change size significantly.) 2. The nature of science 2.1 There’s no such thing as data: the problem of observation So, it looks like we need to think a bit more carefully about what exactly constitutes ‘data’. ‘Data’ is not just there sitting around. It requires interpretation, within the context of a theory which is as explicit as possible. 2. The nature of science 2.2 Idealization: the problem of the ‘real’ world The idea that science studies the real world is another core part of the ‘common sense’ view of science. However, this idea comes with an important set of qualifications. Scientists frequently engage in ‘idealization’. That is, ironically, you need to ‘get away’ from the real world in under to understand how it works. 2. The nature of science 2.2 Idealization: the problem of the ‘real’ world In traditional core natural sciences, with their long tradition, this is understood perfectly well. A chemistry experiment does not consist of digging up rocks from your back garden and mixing it with water from the tap. It consists of mixing some a pure sample of some material (say sodium) and mixing it with water which is pure H2O. Even if nowhere on the actual Earth could you find a ‘pure’ sample of sodium or a ‘pure’ sample of H2O. 2. The nature of science 2.2 Idealization: the problem of the ‘real’ world But if science is about studying the world, it seems like this shouldn’t be legitimate. How can you claim to be studying the world when your experiments contain ‘made up’, ‘unreal’ material that you’ve manufactured in a lab, rather than what’s out there in the world? However, when it comes to chemistry, this is never thought to be a problem. This is because everyone understands that the real world is complicated, and full of interfering factors. If you were to attempt to engage with the world in all its complexity, your investigation would never progress. You’d never get anywhere. 2. The nature of science 2.2 Idealization: the problem of the ‘real’ world What you need to do is try to make the problem simpler. You try to factor out as many interfering factors as you can to try to get to the underlying basic truth about the world. Water that you get out of a tap contains all kinds of stuff – H2O, plus chlorine, plus [insert favourite conspiracy theory here]. If you put sodium into the water and it starts sparking, is that due to the H2O, or the chlorine, or the mind-control drugs? If you dig up a random rock, dump it in pure H2O and the liquid turns blue, which one of the 80 gajillion things that are in a random rock was the cause? You have no way of knowing. 2. The nature of science 2.2 Idealization: the problem of the ‘real’ world So there’s an important sense in which science isn’t about observing the world. It’s about trying ultimately to understand the world, but it sometimes involves dealing with ‘idealized’ versions of it – versions which sometimes incorporate assumptions that you know not to be true. 2. The nature of science 2.3 David Hume and the Problem of Induction Not only does the philosophy of science warn us that we need to think about what ‘observations’ and ‘data’ are, and that we’re not really observing the world. It turns out that the ‘common sense’ version of how theories are built up and justified also isn’t as simple as it seems. 2. The nature of science 2.3 David Hume and the Problem of Induction One of the key things that science does, on the common sense view, is produce laws or generalizations, say like “All copper conducts electricity”. These scientific laws consist of true, certain knowledge about the way the world/universe actually is, and are based on a finite, but appropriately large and varied, number of observations of particular instances of the way the world works. So, given the following set of observations: A bar of copper conducted electricity at time t1 A bar of copper conducted electricity at time t2 A bar of copper conducted electricity at time t3…etc one is entitled to conclude that the following general statement is true: All copper conducts electricity 2. The nature of science 2.3 David Hume and the Problem of Induction However, David Hume (1711-1776) a Scottish philosopher (and economist and historian) who the Stanford Internet Encyclopedia of Philosophy calls “the most important philosopher ever to write in English”, pointed out a big problem. This problem goes by the name of “The Problem of Induction”, though Hume himself didn’t quite put it in those words. 2. The nature of science 2.3 David Hume and the Problem of Induction Inductive reasoning (also called ‘inductive inference’) is “more of the same” reasoning, and this is precisely what the common sense view says science is doing – concluding that a general statement must be true/valid on the basis of (a large number of) relevant individual statements. So, again, given a suitably large and varied number of observations like a bar of copper conducted electricity at time t1, a bar of copper conducted electricity at time t2, etc., scientists have concluded that it is true that all copper conducts electricity. 2. The nature of science 2.3 David Hume and the Problem of Induction But in order to say that it is true that all copper conducts electricity, that is, in order to say that we have a rational basis for the belief that all copper conducts electricity, we need to justify that moment when we go from a finite number of observations to the truth of some general conclusion. We need to be sure that when we use inductive reasoning, we’re using a method of reasoning that’s valid and which leads to the true, certain knowledge that science is supposed to be about. 2. The nature of science 2.3 David Hume and the Problem of Induction But Hume says that the required justification just doesn’t exist. We’re not even entitled to conclude that the next piece of copper will conduct electricity, much less that all pieces of copper will do so. This is because there’s a problem at the heart of inductive reasoning. Inductive reasoning depends crucially on the assumption that the future will be like the past. Or, with respect to the above example, that future copper will be like past copper in terms of conductivity. But this claim (that future copper will resemble past copper) is precisely what we’re trying to establish. And if your argument requires that you assume as part of it the conclusion that you’re trying to prove, the argument is invalid. (This is what ‘circular reasoning’ is (and, technically, also what ‘begging the question’ is).) 2. The nature of science 2.3 David Hume and the Problem of Induction In actual fact, we all know from our daily lives that inductive reasoning isn’t reliable. It isn’t a source of truth or certainty. A famous (if slightly grisly) example of the unreliability of inductive reasoning comes from the British 20th century philosopher Bertrand Russell, and goes by the name of “Russell’s Inductivist Turkey”. Russell’s Inductivist Turkey notes that on January 13th, he was fed at precisely 9AM. He notes that on January 14, he was also fed at precisely 9AM. The same thing happened day after day, and by August or so, the turkey used inductive reasoning to conclude that “I am fed every day at 9AM” is true. Then Christmas Day arrived, and the turkey was not fed at 9AM. 2. The nature of science 2.3 David Hume and the Problem of Induction We use inductive reasoning all the time to get around in the world and to make choices. If I really didn’t think the sun was going to rise tomorrow, I wouldn’t bother doing any work on next week’s lectures. But people use lots of things to get around in the world (superstitions, prejudices, insane hopes that have no chance of being true), and if science is on a par with these, we’re in big trouble. 2. The nature of science 2.4 Why is all of this particularly important for linguists? The scientific study of language in the modern era is a relatively young field (less than 100 years old, as compared with chemistry, which has been studied systematically for over 2500 years). Other natural sciences have long-established practices and methodologies and a rich tradition of investigation. The study of language also can involve a far greater range of objects than many natural sciences. Linguists study everything from texts to introspective intuitions to statistical traits of populations. 2. The nature of science 2.4 Why is all of this particularly important for linguists? It’s vitally important that linguists have a good understanding of what science really is, rather than relying on a ‘common sense’, and wrong, idea of what their work entails. In the weeks to come, we’ll highlight a number of issues and misunderstandings that arise from this problem.