 This e-lecture discusses the dramatic impact computers have had on the discipline of linguistics using a variety of applications as examples. We will look at the following linguistically oriented fields within which computers have been playing an important role over the last two decades. We will deal with linguistics, of course, with the field of artificial intelligence and computational linguistics. We will then look at several applications in these fields and will finally list some aspects that are used in linguistic teaching and research. Now, there are numerous ways in which computers can be used in linguistics and there are various scientific disciplines that are involved in the development of language products. In artificial intelligence, for example, linguistic applications only play a subordinate role. Computational linguistics, in a strict sense, subsumes all those linguistic applications that involve human language processing strategies. Furthermore, there are various language products that lack sophisticated problem-solving strategies but are useful for the linguists in many ways. And last but not least, the computer has become enormously important in linguistic education and research. Let's look at these disciplines and branches in more detail. Now, artificial intelligence, often abbreviated as AI, is the study and design of computer programs that behave intelligently. Thus, artificial intelligence is more than an engineering discipline. It is also subject of scientific investigation. Researchers construct theories about what AI applications should be capable of and test them mathematically and experimentally. Consequently, there are two approaches to artificial intelligence. Artificial intelligence theory and artificial intelligence practice. Let us look at both branches in more detail. Now, over and above, it's engineering aspects. Artificial intelligence is a subject of scientific investigation. For example, it seeks to answer questions such as what adaptations are possible for systems that learn from experience? Or how do systems change in response to new information? Furthermore, what kind of training should a program receive? Well, let us illustrate the theoretical approaches of artificial intelligence. For example, there is a branch dealing with machine vision, which is concerned with the interpretation of complex digital images. Research in artificial intelligence has shown that this is not merely a process of pattern recognition, but an enormously complex process involving several distinct views of the object in question. In machine learning, there are various theoretical approaches to learning. And these involve the finding of learning hypotheses, the modelling of the learner or the setting up of general principles of learning. In medical diagnostics, probabilities are used to infer the most likely disease from a patient's symptoms. The calculations for such inferences are computationally enormously complex. Artificial intelligence wants to find out ways to use such probabilities for diagnostic reasoning. Or take automatic planning. The design of industrial robots makes it desirable that an algorithm never considers the same plan twice and always finds a solution to the plan, if there is one. The finding of algorithms for particular classes of planning problems is an important task of artificial intelligence. Well, and last but not least, we have the field of natural language processing. A field we will say something about later on. Well, and the second approach in artificial intelligence? Well, here artificial intelligence is the basis for a variety of practical systems. Some of these already exist. Some of these are already designed for future use. Here are some examples. The first is referred to as machine translation. Now, machine translation is already there. We use it quite often, don't we? Now here's a sentence I want to translate. Künstliche Intelligenz hat eine praktische und eine theoretische Seite. And I want to translate that into English, and here is a solution. Artificial intelligence has a practical and a theoretical side. We cannot complain about this translation, can we? Well, we will not reveal how this translation is achieved by Google, but this, of course, is an important issue in artificial intelligence practice. Let's return. Now, another practical application in artificial intelligence deals with tutoring, intelligent tutoring. For a long time, artificial intelligence has been interested in machine-based learning. Numerous approaches have been developed. You might have come across some of them. Computer-assisted instruction, CAI. Intelligent computer-assisted instruction, ICAI. Intelligent tutoring systems, ITS, computer-based training, CBT, and so on and so forth. None of these approaches has reached the quality of a human teacher. Nevertheless, a number of important consequences for the learning process have emerged. Or look at traffic control systems, air traffic control systems. Well, due to the popularity of air travel, our skies are becoming increasingly crowded. The same applies to our streets, where traffic congestion is a growing problem. By means of artificially intelligent systems, the arrival and departure of flights, as well as passenger safety, can be maximized and delays can be reduced to a minimum. In standard traffic, AI systems can optimize the use of streets where space is limited. And then you might have come across the field of robotics, where we know that there are various hazardous conditions for humans. For example, toxic waste cleanup, biohazard handling in hospitals, underground mining, space exploration, and so on and so forth. In contrast to humans, intelligent robots are well-equipped to handle such conditions. All these systems require skills we normally associate with natural intelligence. General problem-solving strategies, common-sense knowledge, and a variety of complicated decisions. Let's now look at computational linguistics. Now, computational linguistics can be defined as a special area of artificial intelligence on the one hand. We've already seen this, or as a special branch of linguistics. As an area of artificial intelligence, it has a relatively long history. Going back to 1949, when first approaches towards machine translation were developed by Warren Weaver in the United States. As a branch of linguistics, computational linguistics is much younger. It was not before the beginning of the 1970s when linguistic ideas began to penetrate artificial intelligence. Computational linguistics can be defined as a truly interdisciplinary subject. It is closely connected with linguistics, and linguistics, as you know, contributes to an understanding of the special properties of language data and provides theories of language structure and use. It is also connected, of course, with computer science, which contributes to computational linguistics in many ways. Most importantly, by providing the techniques of software development and maintenance. It is also linked with psychology, which contributes to computational linguistics by discussing the general principles of representing knowledge in the human mind and providing architectures of the human mind. Whereas knowledge representation is supported experimentally. In many ways, the construction of mental models is almost a philosophical enterprise, and philosophy is, of course, also connected with computational linguistics. The philosophical contribution to computational linguistics beyond architectures of the human mind is primarily restricted to the principles of logic. These principles help to formalize linguistic aspects, such as meaning, and constitute the basis for specific programming techniques and languages. Some of you might have come across the programming language prologue. Like artificial intelligence in general, computational linguistics has a theoretical and a practical goal. It uses computers to discover how humans process language, and it enables intelligent computer interfaces for man-machine interaction. There are several computational linguistic applications. For example, machine translation. Well, we've already seen the Google example earlier on, so I don't have to repeat it. Man-machine interfaces. Well, ideally, such interfaces between man and machine use speech in an output. And the output is then some sort of synthetic speech. Here is an example. Well, let's listen. Hello, this is a synthetic computerized voice. My strategy is concatenating very recorded segments of human speech. I'm using a variety of American English. Okay, this is relatively old, but it works. Synthetic speech. Another area or application in computational linguistics deals with information extraction and retrieval. Now here, computers are programmed to analyze the exploding amount of texts available on the web and elsewhere in order to extract selected bits and present them in a structured way or to retrieve documents that contain information of interest to the user. Well, at last but not least, as we've already seen, intelligent tutoring. Now, as I've already briefly mentioned, intelligent tutoring systems attempt to mimic human teaching methods and behaviors using techniques from the field of artificial intelligence. Natural language dialogue capability between learner and program is an important feature of these systems. So here, computational linguistics comes in. The development and theoretical underpinning of all these systems which are experimental rather than operational in many cases is a major goal of computational linguistics. Last but not least, there are further applications of computers, especially in linguistics, without applications without applying general problem-solving strategies or principles of human language processing. The following applications can be defined in this respect. For example, there are search and analysis products. And they are used in a variety of linguistic applications in simple text search. For example, when we want to look up specific word classes or predefined strings in particular texts. Database search systems where we are looking up words in specific, specific linguistic applications such as WordNet for example. Or corpus search where we want to find particular words, sentences, strings in well-known corpora such as the Survey of English Usage or the British National Corpus to name just two of them. These applications may be fully operational without linguistic support, but can be augmented with linguistic routines, for example, with those that generate information about word classes. And then we have the large area of multimedia products. Now multimedia products occur in all branches of linguistics, primarily however in digital learning environments where they can be effectively used in order to model linguistic theories, to exhibit language data, to design linguistic exercises and so on and so forth. Especially in linguistics as a science, computers constitute indispensable tools. Well, and here are the two main applications. For example, in teaching and learning. Many aspects of linguistics can be most efficiently taught using computer support. Just look at the virtual linguistics compass and the thousands of audio examples, videos and simulations, animations, then you will understand what I mean. And in research, well, in research, linguistic research today is impossible without computers. As I said, corpus analysis, text search, sound analysis and sound representation. No serious linguist can publish anything about an exotic language, for example, without providing sound support so that we can look at the primary research data. This is for example why we have created the language index and this is what it looks like. Well, here is our new language index with currently just a thousand sound examples, but pretty soon there will be many more and this is what you can do with it. For example, you're looking for a language, let's say you want an example from French Creole, spoken in Dominica. Well, here we are. We immediately find a speaker and the speaker can welcome us, of course. Welcome everyone to the linguistics website. My name is Jeanette Lando. I speak the language of Dominique. Or if you want information about the morphology and syntax. Well, you can find it in the language index. How could you present such a language without computer support? Well, let's summarize. This lecture has shown the dramatic impact computers have had on the discipline of linguistics. It has also illustrated the influence of neighboring disciplines such as psychology or philosophy. Furthermore, we have seen that there are straightforward computational linguistic applications as well as a large number of applications that support linguistic teaching and linguistic research. One aspect is common to all of these. The role of the computer in linguistics will increase rather than decrease. In what way will be discussed in further e-lectures about linguistic engineering? So, see you there.