下载文档原格式
Unit 2 Functions of languageOverviewTo further explore the nature of language, we naturally come to the question of what language is for. This unit deepens our understanding of the basics of language by offering refreshing ideas about language functions, which pave the way for the investigation of language in use, language evolution, language acquisition, and other important issues in linguistics.Text ALanguage is used for so many purposes (e.g., expressing feelings, providing information, and conducting social talking, to name just a few) that it is difficult for us to identify which one is its most basic (perhaps also its) original use. However, a careful examination of what language is good/bad at today can hopefully provide some useful clues to its basic function(s). Contrary to the mainstream assumptions, it has been observed that among its multiple purposes, language is good at interaction and persuasion but poor at information exchange.Text BThe hammer (with an iron head and a handle) is shaped to fulfill its functions of pounding nails into and removing them out of wood. Likewise, the form of language is also motivated by its basic functions. As a form-function composite, a linguistic unit consists of two parts, the signifier and the signified. The bond between the two parts is both arbitrary and motivated. It is arbitrary because there is no inherent connection between the form of a sign and its meaning; it is motivated in the sense that the bond between the signifier and the signified is intentional instead of rigid or random.The two texts approach the same issue from different perspectives. Text A discusses how the basic functions of a language can be traced in the development of language, while Text B focuses on the intriguing relationship between the form and the functions of language.Teaching objectivesThis unit is designed to help students develop their reading skills, communicative competence, critical thinking, intercultural competence and abilities of autonomous learning in the following aspects.Reading skills:Distinguish between general and specific statementsScan for a global understanding of the textRead charts, graphs and tables to organize and interpret informationCommunicative competence:Illustrate your points with appropriate examplesUse figures of speech to help Illustrate your pointsDifferentiate informative and affective uses of languageCritical thinking:Evaluate the logical strengths of the author’s evidence and supportOrganize and present your reasoning using diagramsIdentify and critique assumptions behind statementsIntercultural competenceAppreciate linguistic and cultural diversityInterpret language differences culturallyHave curiosity in and be ready to explore other languages and culturesTeaching strategiesThe functions of language are an interesting yet rather difficult topic. To arouse students’ interest, the teacher can start with a direct question “What do you think is the function of language?” or a more interesting one “In what way is language similar to a hammer?”The questions in Critical reading and Intercultural reflection are mostly mini-research projects. To guarantee quality answers from students and to improve the classroom efficiency, it is recommended to divide the class into groups and assign different tasks to them in advance. Each group is required to do in-depth researches on a few of the questions.Further readings on functions of language:Jakobson, Roman. (1960). Closing statement: linguistics and poetics. In Thomas A. Sebeok (ed.). Style in Language, Cambridge, MA.: The MIT Press. 350-377. Halliday, M.A.K. (1973). Explorations in the Functions of Language. London: Edward Arnold.Coq, John P. (1955). The Function of Language. The Modern Language Journal, 39(4): 177-180.Jackson, Howard & Peter Stockwell. (2011). An introduction to the nature and functions of language. London & New York : Continuum.Preparatory work(1)Main publications:New media language. London and New York: Routledge, 2003.Words in the mind: An introduction to the mental lexicon. 3rd edition. Oxford and New York: Basil Blackwell, 2003.The Articulate Mammal: An Introduction to Psycholinguistics. 4th edition. London and New York: Routledge, 1998.The seeds of speech: Language origin and evolution. Cambridge, New York, Melbourne: Cambridge University Press, 1996.Main research interests:1)(Socio-)historical linguisticsThe description, implementation and causation of language change, with particular reference to current changes.2)Language and mindLanguage acquisition, speech comprehension, speech production, with particular reference to lexical storage and retrieval.3)Language and the mediaThe language used by the media, and the effect of the media on language, with particular reference to language change, language and power, the relationship of media language to language in literature.(2)The whole book consists of four parts, namely, Puzzles, Origin, Evolution, and Diffusion. Text A “What is Language for” is from the first part Puzzles. The other three questions addressed in this part are: How did language begin? Why do languages differ so much? and Is language an independent skill?Part 2, 'Origin', explores how language probably originated. Chapter 5 outlines the 'East Side story', which suggests that humans separated from apes when they were stranded on the east side of Africa, after the Great Rift Valley split the terrain. Humans were forced to live on their wits in a harsh landscape, and began to develop language. Chapter 6 examines the prerequisites for language, which are shared with our ape cousins: first, friendly involvement with others and a predisposition for grooming; second, an aptitude for tactical deception, or lying, which depends on 'a theory of mind', and an ability to understand the intentions of others. Chapter 7 considers the basic requirements for speech, many of which are also present in other primates. For example, sound-receiving mechanisms are shared with apes, yet sound-producing ones differ, perhaps because of our upright posture, which in the long run enabled humans to produce a range of finely tuned sounds. Chapter 8 notes that ontogeny, the development of the individual, only sometimes correlates with phylogeny, the development of the species. Two ways in which these coincided were the lowering of the larynx and the development of the 'naming insight', an understanding of the power of naming.Part 3, 'Evolution', examines how language might have evolved. Chapter 9 looks at ways in which words could have been combined. At first, many sequences were possibly repetitive and inconsistent. But gradually, strong preferences may have become rules. These preferences were based on pre-linguistic 'mind-sets'. Chapter 10 looks at how language expanded and how it made use of the human body and its location in space for extending word meanings. The evolution of different parts of speech occurred probably via reanalysis: adjectives and prepositions both grew out of reinterpretation of nouns and verbs. Chapter 11 considers attachments to verbs. Verbs acquire attachments via grammaticalization, in which a full lexical item developed into a grammatical marker. Chapter 12 examines generativity, the use of finite resources to produce an infinite variety of sentences. Such structures arose from reanalysis of existing structures.Part 4, 'Diffusion', considers the spread of language over the world, and discusses why languages have not become unlearnably different from one another. Chapter 13 outlines the route taken as humans moved out of Africa, and considered the possibility of reconstructing glimpses of language as it might have been more than 30,000 years ago. Chapter 14 examines the difficulties and frustrations of hunting for language universals. Chapter 15 points out that looking for constraints, things that languages don't do might be more enlightening. Implicational links, it notes, are important for keeping language in check. Such links are partially due to processing needs, but partly also to the overall structure of the system. Chapter 16 outlines the different components within the language system, and emphasizes that these interacted with the usage of the system in a complex way.(3)John Locke, widely known as the Father of Classical Liberalism, was an English philosopher and physician. He was also memorized as one of the first British empiricists. His work greatly affected the development of epistemology and political philosophy. His contributions to classical republicanism and liberal theory are reflected in the United States Declaration of Independence.Major works:A Letter Concerning Toleration (1689)Two Treatises of Government (1689)An Essay Concerning Human Understanding (1690)Some Thoughts Concerning Education (1693)(4)Lord Byron’s (1788-1824) two best known long narrative poems are Don Juan and Childe Harold's Pilgrimage. His other major poems include: Prometheus, She walks in Beauty, When we Two parted, Darkness, and And Thou art Dead, as Young and Fair. (5)Bronislaw Malinowski (1884-1942) is remembered as the father of the functionalistschool of anthropology and for his role in developing the methods and the primacy of anthropological fieldwork as well. He first rose to prominent notice through his studies of Pacific Islanders, especially those conducted among the Trobriand Islanders whose marriage, trade and religious customs he studied extensively. His best known works include Argonauts of the Western Pacific (1922),Crime and Custom in Savage Society (1926), The Sexual Life of Savages in North-Western Melanesia (1929), and the posthumously published Magic, Science, and Religion and Other Essays (1948). Malinowski helped develop the field of anthropology from a primarily evolutionary focus into sociological and psychological enquiries. Some of the noteworthy byproducts of his fieldwork in this direction include various evidence that debunked the Freudian notion of a universal Oedipal Complex and evidence that showed that so-called primitive peoples are capable of the same types and levels of cognitive reasoning as those from more "advanced" societies. Malinowski's ideas and methodologies came to be widely embraced by the Boasian school of American Anthropology, making him one of the most influential anthropologists of the 20th century.Source: /people/320/000099023/For more information about Malinowski, please refer to: Murdock, George(1943). Bronislaw Malinowski. American Anthropologist, 45:441-451./committees/commissions/centennial/history/095malobit.pdf Critical readingI. Understanding the text1. OutliningThesis: Among its multiple purposes, language is good at interaction and persuasion but poor at information talking.We would like to emphasize that what we have just provided is not the only answer to the outlining task. To start with, we can divide the text in different ways. For example, we can follow the traditional trichotomy and dividing it into introduction, main body and conclusion; we can also divide part III and part IV into further sections. The main idea of each part varies accordingly.2. Comprehension check(1)The transfer of information is not the only purpose of using language. In greetingsand some pointless chitchat, communicators use language primarily, if not exclusively, for constructing or maintaining certain social relationship. Even when information is the major concern, the speaker and hearer must take into consideration some other factors, such as politeness and aesthetics (cf. Paras. 22-22).(2)In Paragraph 6, Aitchison divides the question “what is language for” into twosub-questions: “For what purpose did language develop?” and “For what purpose is language used nowadays?” Since there are so many purposes of using language and the original one is difficult to identify, she argues in paragraph 8 that we can find clues by looking at what language is good at and what it finds difficult to express. The rest of the text (paras.9-28) provides discussion about and answers to these two questions.(3)According to Aitchison, the list of language functions in paragraph 7 is not exhaustive,and it is not clear which one is the most basic. Aitchison discusses in some detail thefollowing four functions, providing information, expressing feelings, influencing others and social talking, which are roughly organized in the order of importance in the traditional view.(4)Aitchison suggests that the early functions of language can be traced in the way weuse language today to some extent. The assumption behind is that the origin of language is accountable in its early function(s) and that the early function(s) must be reflected in what language is good at today. In other words, if language was created to perform a particular function, it must still be good at it nowadays.II. Evaluation and exploration1.Evaluating the text(1)Figure drawing(2)ExemplifyingAn utterance may serve more than one purpose simultaneously. Donking is used metalinguistically in example (6), but the whole sentence “Donking isn’t a word” is informative. We can provide information, express our feelings and initiate social talking by asking questions or giving commands.2.Exploring beyond the text(1)Some scholars believe that language facilitates thinking and that our thinking wouldbe impossible without an inner language. In many cultures language is also used as a symbol of magic or as something that carries mysterious power (e.g., religious Taoism, couplets for Spring Festival). Different functions of language are not equally important. For example, the functions of communicating and of facilitating thinking are more fundamental while those of religious use and word play are more derivative.(2)Generally speaking, there are two different views on the relationship betweenlanguage and thinking. Some scholars claim that language restricts thinking.According to this view, people perceive the world through the language that they speak. It follows that people speaking different languages experience different worlds, just like people seeing different things with different eyes. Others believe that language and thinking are separate and should not be equated with each other. For example, even if a language does not have the word for chartreuse, people speaking the language can still perceive this color, think about it and even talk about it,using not a single word but a kind of paraphrase.(3)Some utterances may basically serve only one function. For example, people greeteach other saying “Hi!” or “Morning!” to neighbors to maintain social connections; inchurch, the priest pr eaches a sermon to call for piety to the Lord. But more often than not an utterance and its context produce some “side effects” and serve different functions simultaneously. In saying “It’s gonna rain. You’d better take an umbrella.”to a lady, the speaker not only provides information about the weather, but also shows his/her concern toward the addressee and enhances the solidarity.(4)Language changes for different reasons, e.g., language users’ a version to cliché andpreference to creativity, language contact, language planning and so on. While some changes may not be directly related to functions of language, others are indeed motivated by certain functions of language. For example, neologisms related to science and technology emerge primarily because there is a need to talk about such new things. In this case the function of providing information requires the creation of some new words.(5) The best-known examples for “phatic communion” in tradi tional Chinese society areprobably greetings such as “吃了吗？(have you had your meal?)”and “去哪儿啊？(where are you going?)”. Neither is considered imposing or offensive because in traditional Chinese society, it is more important to show mutual concern than respect for privacy. The British follow a different tradition. Brown and Levinson (1987) recognize two sides of face, a positive one and a negative one. Positive face refers to “the desire to be appreciated and approved” and negative face “the freedom of act ion and freedom from imposition”. For Chinese people (especially in traditional society), negative face is not as important as positive face. The British, on the other hand, value negative face more than positive face.Language EnhancementI. Words and phrases1. Word formationPart of speech:Nouns: Taking-off, air-traffic, take-off-point, London-York, two-thirds, real-life, non-reality, self-repetition, other-repetitionAdjectives: non-existentAdverbs: half-wayStructure:Noun-noun: air-traffic, London-York, self-repetition, other-repetitionAdjective-noun: real-lifeCardinal-ordinal numeral: two-thirdsPrefix-noun: non-realityGerund-adverb: Taking-offVerb-adverb-noun: take-off-point2. Articles and prepositions(1)/(2)/(3) a, the, with(4) /, the(5) /, the, the, on, a3.Verbs and phrases(1)convey, handle(2)convey(3)transfer(4)coincide(5)collide(6)date back to(7)originate fromII. Sentences and discourse1. Paraphrasing(1) Even when language is used simply to transfer information, the accuracy of the information transferred is still unlikely to reach 100%.(2) Less affirmative claims about the purpose of language may be at least superficially more acceptable: when early people found that facial expressions and body movements could not fully express themselves, they invented language to communicate their thoughts.(3) So long as the speaker is telling the truth, language is reasonably good at transferring sim ple pieces of factual information, such as “Bob is Petronella’s cousin.”(4) Even in cultures where lying is officially discouraged, people are still unwilling to tell the whole truth. A government official invented the phrase “being economic with the tru th” to deny that he was lying.(5) According to George Orwell, political language is designed to beautify horrible things and to tell lies.2. Translation(1) 英国哲学家约翰·洛克在其颇有影响的著作《人类理解论》(1690)中指出，语言是一个伟大的传输渠道，人类通过这一渠道相互传达各自的发现、推理和知识。
logic and philosophy1.Leibniz was the real implementer of the modern universal language program. He not only reexpressed the three major formal logic in a symbolic way, but also put forward the seven axioms of logical calculus, thus starting the work of logical mathematics. He made a comprehensive investigation of the relationship between logic and metaphysics after Aristotle, and proposed for the first time forward the fundamental consistency of the two. His discussion on concepts, definitions and propositions has an incentive effect on logic. His distinction between analytical propositions and comprehensive propositions has become an important ideological resource for Kant's philosophy and Husserl phenomenology.The significance of logic to metaphysics: Since Aristotle, logic has been closely related to metaphysics and epistemology. Metaphysics has always been regarded as the knowledge "about existence as existence", as the pursuit of the first principles of the world, while logic has always been regarded as the form and law of thinking. The epistemological transition realized by modern philosophy not only provides new opportunities for the reconstruction of the internal connection between logic and metaphysics, but also expands the theoretical domain and vision of both. Among the seventeenth-century philosophers, Leibniz was the most clear and most fully expressed the basic ideas of logical philosophy. In his place, logic was both the great tool of reason, the fundamental expression of philosophical truth, and the fundamental principle of philosophical research, because, in his opinion, "everything created through reason can be created by perfect logical rules". Leibniz tried to base philosophy in the traditional sense firmly by establishing the value of logical reason, because he found that philosophy lacks a clarity and certainty. Therefore, he hopes to transform philosophy logically so as to make philosophical concepts, propositions and reasoning realistic. In the New Theory of Human Reason, he agrees with the view: " The function of philosophy is to create some words, in order to give people definite concepts, and to express the definite truth in general propositions.”2.According to Leibniz's idea of classification, there are two main treatments of the truth of all doctrines, each with its own weight and each having its own value and significance, but the best way is to combine them, because they complement and complement each other. These two methods are integrated (also known as theoretical) and analytical (also known as practical). The comprehensive method or theoretical method is to arrange the truth in the order of the proof. Like a mathematical proof, put each proposition after the proposition taken as a premise. In this way, all the propositions representing the truth will present a progressive logical relationship. Analytic or practical method starts with the purpose of man, starts with good, from the highest point of good, human happiness, and then transition to the special means of achieving good (or avoiding the opposite of good, evil). In this sense, analytical methods are designed to transition from purpose to means, from abstract entry to special, or to decline from general to individual. In addition to the above two treatments, Leibniz believes, we can also add a third method, namely, a method of arranging the truth by a noun, which is actually an indexing method, which Leibniz used for book classification and cataloging. Leibniz said that the third method is equivalent to the ancient logic method, because it dealswith knowledge and truth according to certain categories of logic, which involves both the understanding of the nature of species and genera, as well as the definition of the logical extension and connotation of categories. The above classification is consistent with the scientific classification of the ancient Greeks. Because the ancient Greeks divided philosophy or science into three categories: theoretical, practical, and ethical knowledge. Theoretical knowledge is equivalent to what Leibniz has here to as comprehensive law, practical knowledge is equivalent to analytical law, and the method of arranging truth by noun is equivalent to logic.As Leibniz's mind matured, he attached more attention to logic. He said: " As for logic, which is the teaching and connection of thought, I see no reason to blame. On the contrary, the lack of logic makes people wrong."He not only greatly expanded the scope of logic, but also tried to reconstruct metaphysics from the logical analysis of propositions. To this end, he needs both to revisit and establish the logical premise of metaphysics, and also to establish a reliable logical method in order to discover and express the definite truth. According to Russell, there are five main premises of Leibniz's philosophy:(1) Each proposition has a primary term and a predicate term;(2) A primary term may have several predicates about the properties that exist at different times;(3) All true propositions that constantly claim existence at a certain time are inevitable and analytical, while those that assert existence at a certain time are accidental, with the latter relying on the ultimate cause;(4) The self is an entity;(5) Perception produces knowledge about the outside world, that is, about myself and about existing things outside of my state.Clearly, the first three of these five premises are related to logic. The fourth premise is the basis of Leibniz's epistemology, and, in a sense, it is also the basis of his metaphysics. But this premise is also indirectly related to logic. As Russell said, " The entity concepts, as we will understand, are derived from the logical concepts of both the principal terms and the predicate terms."If the concept of the entity is one of the most basic concepts of Leibniz's metaphysics, then the logical concept of the subject and predicate term also establishes its metaphysics independently. For this reason, Russell asserted that " Leibniz's metaphysics comes from his subject and predicate logic."He even asserts," Leibniz's philosophy stems almost entirely from his logic.”。
第一章绪论principles and theories to language Modern linguistics regards the imagined matters in the past1/ What is linguistics teaching and learning especially the spoken form of language as primary present or future or in far-away什么是语言学？teaching of foreign and second but not the written form. Reasons places.Linguistics is generally defined as language. are: 1. Speech precedes writing 2. Cultural transmission语言的文化the scientific study of language. It Other related branches include There are still many languages that 传递性studies not any particular language anthropological linguistics （人类have only the spoken form 3. In While human capacity for languagebut languages in general. 语言学）neurological linguistics terms of function the spoken has a genetic basis i.e. we were2/ The scope of linguistics （神经语言学）mathematical language is used for a wider range born with the ability to acquire语言学的研究范畴linguistics （数字语言学）and of purposes than the written and language the details of anyThe study of language as a whole is computational linguistics. （计算机carries a larger load of language are not geneticallyoften called general linguistics. 普语言学）communication than the written. transmitted but instead have to be通语言学3/ Some important distinctions in Langue and parole 语言和言语taught and learned anew. ThisThe study of sounds which are linguistics The Swiss linguist F. de Saussure indicates that language is culturallyused in linguistic communication is 语言学研究中的几对基本概念made the distinction between langue transmitted. It is passed down fromcalled phonetics. （语音学）Prescriptive and descriptive 规定and parole early 20th century. one generation to the next throughThe study of how sounds are put与描写Langue refers to the abstract teaching and learning rather thantogether and used in communication If a linguistic study describes and linguistic system shared by all the by instinct.is called phonology. （音系学）analyzes the language people members of a speech community Chapter 3 Morphology 形态学The study of the way in which actually use it is said to be and parole refers to the realization Definition 定义morphemes are arranged to form descriptive if it aims to lay down of langue in actual use. Saussure Morphology is a branch of grammarwords are called morphology. （形rules to tell people what they should made the distinction in order to which studies the internal structure态学）say and what they should not say it single out one aspect of language of words and the rules by whichThe study of how morphemes and is said to be prescriptive. for serious study. He believes what words are formed.words are combined to form Modern linguistics differs from linguists should do is to abstract 形态学是语法的一个分支，研sentences is called syntax. （句法traditional grammar. Traditional langue from parole to discover the 究词的内部结构和构词规则。
一、Directions: Please define the following terms.1.minimal pair testwords that differ in only one soundThey differ in meaning, they differ only in one sound segment, the different sounds occur in the same environmentExample: beat, bit They form a minimal pairSo /ea/ and /i/ are different sounds in EnglishThey are different phonemes2.the Sapir-Whorf Hypothesislinguistic determinism (语言决定论) -Language determines thought.and linguistic relativity (语言相对论)-There is no limit to the structural diversity of languages.3.BehaviorismBehaviorism in linguistics holds the view that Children learn language through a chain of stimulus-response-reinforcement (刺激—反应—强化), and adults‟ use of language is also a process of stimulus-response.4.discovery proceduresA grammar is discovered through the performing of certain operations on a corpus of data5.Universal GrammarUG consists of a set of innate grammatical principles.Each principle is associated with a number of parameters.6.Systemic GrammarIt aims to explain the internal relations in language as a system network, or meaning potential.7.Ideational MetafunctionThe Ideational Function (Experiential and Logical) is to convey new information, to communicate a content that is unknown to the hearer. It is a meaning potential.It mainly consists of “transitivity” and “voice”. This function not only specifies the available options in meaning but also determines the nature of their structural realisations. For example, “John built a new house” can be analysed as a configuration of the functions (功能配置):Actor: JohnProcess: Material: Creation: builtGoal: Affected: a new house8.Interpersonal MetafunctionThe INTERPERSONAL FUNCTION embodies all uses of language to express social and personal relations. This includes the various ways the speaker enters a speech situation and performs a speech act.9.basic speech rolesThe most fundamental types of speech role are just two: (i) giving, and (ii) demanding.Cutting across this basic distinction between giving and demanding is another distinction that relates to the nature of the commodity being exchanged. This may be either (a) goods-&-services or (b) information.10.finite verbal operatorsFiniteness is thus expressed by means of a verbal operator which is either temporal or modal.11.Textual MetafunctionThe textual metafunction enables the realization of the relation between language and context, making the language user produce a text which matches the situation.It refers to the fact that language has mechanisms to make any stretch of spoken or written discourse into coherent and unified texts and make a living passage different from a random list of sentences.It is realized by thematic structure, information structure and cohesion.12.theme and rhemeThe Theme is the element which serves as the point of departure of the message.The remainder of the message, the part in which the Theme is developed, is called the Rheme.As a message structure, a clause consists of a Theme accompanied by a Rheme.The Theme is the first constituent of the clause. All the rest of the clause is simply labelled the Rheme13.experientialismExperientialism assumes that the external reality is constrained by our uniquely human experience.The parts of this external reality to which we have access are largely constrained by the ecological niche we have adapted to and the nature of our embodiment. In other words, language does not directly reflect the world. Rather, it reflects our unique human construal of the world: our …world view‟ as it appears to us through the lens of our embodiment.This view of reality has been termed experientialism or experiential realism by cognitive linguists George Lakoff and Mark Johnson. Experiential realism acknowledges that there is an external reality that is reflected by concepts and by language. However, this reality is mediated by our uniquely human experience which constrains the nature of this reality …for us‟.14.image schemataAn image schema is a recurring structure within our cognitive processes which establishes patterns of understanding and reasoning. Image schemas are formed from our bodily interactions, from linguistic experience, and from historical context.15.prototype theoryPrototype theory is a mode of graded categorization in cognitive science, where some members of a category are more central than others. For example, when asked to give an example of the concept furniture, chair is more frequently cited than, say, stool. Prototype theory has also been applied in linguistics, as part of the mapping from phonological structure to semantics.二、Directions: Please answer the following questions.1.Why is Saussure called “one of the founders of structural linguistics and “father ofmodern linguistics”?He helped to set the study of human behavior on a new footing (basis).He helped to promote semiology.He clarified the formal strategies of Modernist thoughts.He attached importance to the study of the intimate relation between language and human mind.2.W hat are the similarities and differences between Saussure‟s langue and parole andChomsky‟s competence and performance?The similarities (1) language and competence mainly concerns the user‟s underlying knowledge; parole and performance concerns the actual phenomena (2) language and competence are abstract; parole and performance are concrete.The differences (1) according to Saussure, language is a mere systematic inventoryof items; according to Chomsky, competence should refer to the underlying competence as a system of generative processes (2)According to Saussure, language mainly base on sociology, in separating language from parole, we separate social from individual; according to Chomsky, competence was restricted to a knowledge of grammar.3.What is the conflict between descriptive adequacy and explanatory adequacy?A nd what is Chomsky‟s solution to this conflict?a theory of grammar: descriptively adequateshould adequately describe thegrammatical dada of a language.should not just focus on a fragment of a language.a theory of grammar: explanatorily adequateshouldexplain the general form of language.should choose among alternative descriptively-adequate grammars.should essentially be about how a child acquires a grammar.A theory of grammar should be both descriptively and explanatorily adequate.But there is a conflict:To achieve DA, the grammar must be very detailed.To achieve EA, the grammar must be very simple. (think why?)because the child can learn a language very easily on very little language exposure.Chomsky‟s solution:construct a simple UGletindividual grammars be derivable from UG4.What are Chomsky‟s contribution s to the linguistic revolution?Chomsky‟s contribution to t he linguistic revolution is that he showed the world a totally new way of looking at language and at human nature, particularly the human mind. Chomsky challenged behaviorism and empiricism because he believes that language is innate.Rationalism (vs. empiricism in philosophy)Empiricist evidence is often unreliable.Innateness (vs. behaviorism in psychology)Children can acquire a complicated language on the basis of very limited exposureto speech.This indicates that UG is innate faculty.5.How to compare and contrast Generative Linguistics and Systemic-FunctionalLinguistics from perspectives of epistemology, theoretical basis, research tasks and methodology?6.How many process types are there in the transitivity system? Please illustrate eachtype by a proper example.Six. Material Processes, Mental Processes, Relational Processes, Behavioural Processes, Verbal Processes, Existential ProcessesThe typical types of outer experience are actions, goings-on and events: actions happen, people act on other people or things, or make things happen. This type of process is called Material Processes.The inner experience is that in our consciousness or imagination. You may react on it, think about it, or perceive it. This type of process is called Mental Processes.Then there is a third type of process: we learn to generalize, to relate one fragment of experience to another. It does this by classifying or identifying. This kind of process is called Relational Processes.These three processes are called major processes. Related to them are three minor processes: each one lies at the boundary between two processes of the three. Not so clearly set apart, they share some features of each, and finally acquire the character of their own.On the borderline between material and mental are the Behavioural Processes: those that represent outer manifestations of inner workings, the acting out of processes of consciousness and physiological states.On the borderline of mental and relational is the category of Verbal Processes: symbolic relationships constructed in human consciousness and enacted in the form of language.Then on the borderline between the relational process and the material process are Existential Processes, by which phenomena of all kinds are recognized to be or to exist.7.What is a multiple Theme to be contrasted with a simple Theme? What is a markedTheme to be contrasted with an unmarked Theme? Please illustrate them with proper examples.Conjunctions in ThemeConjunctive and modal Adjuncts in ThemeTextual, interpersonal and experiential elements in ThemeInterrogatives as multiple Themes8.What are the similarities and differences between conceptual metaphor andconceptual metonymy?Metaphor and metonymy are viewed as phenomena fundamental to the structure of the conceptual system rather than superficial linguistic …devices‟.Conceptual metaphor (概念隐喻) maps structure from one conceptual domain onto another, while metonomy highlights an entity by referring to another entity within the same domain.隐喻就是把一个领域的概念投射到另一个领域，或者说从一个认知域（来源域）投射到另一个认知域（目标域）。
. Introduction1. What is LanguageLanguage is a system of arbitrary vocal symbols used for human communication.2. What is Linguistics(语言学)Linguistics is the scientific study of language.3.Some Basic Distinctions(区分) in Linguistics3.1 Speech and WritingOne general principle(原则) of linguistic analysis is the primacy of speech over writing. Writing gives language new scope(范畴) and uses that speech does not have.3.2 Descriptive(描述性) or Prescriptive(说明性)A linguistic study is descriptive if it describes and analyses facts observed; it is prescriptive if it tries to lay down rules for "correct" behavior.3.3 Synchronic(共时) and Diachronic(历时) StudiesThe description of a language at some point in time is a synchronic study and The description of a language as it changes through time is a diachronic study.3.4 Langue(语言) and Parole(言语)This is a distinction made by the Swiss linguist F.De Saussure (索绪尔）early last century. langue refers to the abstract linguistic system shared by all the members of a speech community and parole refers to the actualized(实际的) language, or realization of langue.3.5 Competence(能力)and Performance(行为)Competence is the ideal language user's knowledge of the rules of his language. Performance is the actual realization of this knowledge in utterances(发声).4.The Scope of LinguisticsGeneral linguistics is the study of language as a whole.Phonetics(语音学) is the branch of linguistics which studies the characteristics of speech sounds and provides methods for their description, classification and transcription.Phonology(音韵学) is the branch of linguistics which studies the sound patterns of languages.Morphology(词法) is the branch of linguistics which studies the form of words.Syntax(句法) is the branch of linguistics which studies the rules governing the combination of words into sentences.Semantics(语义学) is the branch of linguistics which studies the meaning of language.Applied linguistics(应用语言学) is the study of the teaching of foreign and second languages. Sociolinguistics is the study of the relationship between language and society.Psycholinguistics is the study of the relationship between language and the mind.Historical Linguistics(历史语言学) is the study of language changes.Anthropological linguistics(人文语言学) uses the theories and methods of anthropology to study language variation and language use in relation to the cultural patterns and beliefs of man.Neurolinguistics(神经语言学) studies the neurological basis of language development and use in human beings.Mathematical linguistics(数学语言学) studies the mathematical features of language, often employing models and concepts of mathematics.Computational linguistics(计算语言学) is an approach to linguistics in which mathematical techniques and concepts(概念) are applied, often with the aid of a computer.II. Phonetics(语音学）1. scope of phoneticsSpeech sounds may be studied from different angles, thus we have at least three branches of phonetics: Articulatory phonetics(发音语音学）we may examine the way in which a speech sound is produced to discover which vocal organs are involved and how they coordinate(协调）in the process.Auditory phonetics （听觉语音学）we may look into the impression a speaker makes on the hearer as mediated(调节）by the ear, the auditory nerve(神经）and the brain.Acoustic phonetics (声学语音学）we study the physical properties of speech sounds, as transmitted（传送）between mouth and ear.2. The vocal organsThe vocal organs may be viewed as consisting of three parts, the initiator of the air-stream,(气流发生器官）the producer of voice（声音发生器官）and the resonating cavities.（声音共振器官）3. Consonants(辅音）Places of articulation(发音部位）: bilabial,(双唇）Labiodentals,（唇齿）dental,（齿）alveolar,（齿龈）retroflex,（卷舌）palate-alveolar,（上齿龈）palatal,（上颚）velar,（软腭）uvular,（小舌）glottal(声门）Manners of articulation: plosive,（暴破）nasal,（鼻音）trill,（颤音）lateral,（边音）fricative,（摩擦）approximant,（近似音）affricate（破擦）4. Vowels (元音）The classification of vowels: the height of tongue raising (high, mid, low), the position of the highest part of the tongue(front, central, back), and the degree of lip rounding(rounded, unrounded)III. Phonology（音韵学）1. phonemes(音素）：a distinctive(有区别的）sound in a language.2. Allophones(音位变体）：The nondistinctive sounds are members of the same phoneme.3. Minimal pairs(最小对立体）：word forms which differ from each other only by one sound.4. Free variation (自由变异）：If two sounds occurring in the same environment(环境）, they does not produce a different word form, but merely a different pronunciation of the same word.5. Complementary distribution(补充分类）：Not all the speech sounds occur in the same environment. When two sounds never occur in the same environment6.Suprasegmental phonology(超音段音位）：the study of phonological properties(性质）of units lager than the segment-phoneme. They are syllable(音节），stress,（重音）word stress, sentence stress. pitch （音调）and intonation（语调）.IV. Morphology（词法）1. inflection(构形法）：the grammatical relationships through the addition of inflectional affixes.(屈折词缀）2. Word-formation(构词）：the processes(过程）of word variations signaling lexical relationships.（表明词法关系）They are compound(合成）and derivation (派生）.3. Morpheme(词素）：the smallest unit in terms of relationship between expression and content.4. Allomorph(同质异象变体）：some morphemes have considerable variation, for instance, alternate shapes or phonetic forms.5. Types of morphemes: They are roots,(词根）affix（词缀）and stem（词干）.6. Lexicon(语言词汇）：in its most general sense, is synonymous with vocabulary.7. Closed-class words(封闭性）and open-class words(开放性）：the former whose membership is fixed or limited and the latter whose membership is in principle(实际上）indefinite or unlimited.8. Word class(词性）：It displays a wider range of more precisely defined classes.9. Lexeme(词位）：the smallest unit in the meaning system of a language that can be distinguished from other smaller units.10. Idiom(习语，成语）：Most phrasal lexemes are idioms. It is especially true for a sequence of words(词序）which is semantically（语义上）and often syntactically(句法上）restricted.（限制）11. Collocation(搭配）：the habitual(习惯的）co-occurrences （同时出现）of individual lexical items. V. Syntax (句法）1. Positional relation or word order（词序）：the sequential(顺序）arrangement of words in a language.2. Construction or constituent (句子结构）：the overall process of internal (内部）organization of a grammatical unit .3. Syntactic function(句法功能）：the relationship between a linguistic form and other parts of the linguistic pattern in which it is used. The names of functions are expressed in terms of subjects, objects, predicates, modifiers,(修饰语）complements（补语）, etc.4. Category(范畴）：It refers to classes and functions in its narrow sense, e.g. noun, verb, subject, predicate, noun phrase, verb phrase, etc. The categories of the noun include number, gender, case and countability.5. Phrase: a single element of structure containing more than one word, and lacking the subject-predicate structure typical of clause.6. Clause: a group of words with its own subject and predicate, if it is included in a larger sentence.7. Sentence: It is the minimum part of language that expresses a complete thought.VI. Semantics1. Conceptualism or mentalism (概念主义）：Following F. De Saussure(索学尔）'s "sign" theory, the linguistic sign is said to consist of a signifier （所指）and signified（被指）, i.e., a sound image and a concept, liked bya psychological（心理的）"associative" bond.（相关联系）2. Mechanism(机械主义）：Some linguists, Bloomfield,(布鲁费尔德）for example, turned to science to counter（反）-act the precious theories and this leads to what call the mechanistic approach（方法）. The nature of this theory has nothing to do with the scientific study of mental phenomena.（智力现象）3. Contextualism （语境主义）：It is based on the presumption(假定）that one can derive meaning from or reduce it to observable context.4. Behaviorism (行为主义）：Behaviourists attempt to define (定义）the meaning of a language form as "the situation(情景）in which the speaker utters（说话）it and the response（反应）it calls forth in the hearer."5. functionalism (功能主义）：Functionalists as represented (代表）by the Prague school（布拉格学派）linguists and neo-Firthian （新弗斯）linguists, approach the problem from an entirely new orientation（方法）. They argue（争辩）that meaning could only be interpreted（解释）from its use or function in social life.6. Sense relationships: While reference deals with the relationship between the linguistic elements, words, sentences, etc.,and the non-linguistic world of experience, sense relates to the complex system of relationships that hold between the linguistic elements themselves. They include synonymy(同义词），antonymy(反义词）,hyponymy(下层次）Polysemy(一词多义）and Homonymy (同音异义词）7. Semantic analysis: It includes 1) componential（成分）analysis which defines the meaning of a lexical element in terms of semantic components.(意义成分）2) predication (表述）analysis in which the meaning of a sentence is not merely the sum of the meanings of the words which compose it. 3) relational components in which the semantic analysis of some words presents a complicated picture, because they show relations between two and perhaps more terms.VII. Language Variation (语言变化）1. Lexical change(词汇的变化）：changes in lexis.2. Invention: (新造词）new entities.3. Compounding:(合成词）New words are sometimes constructed by combining two old words.4. Blending: (混合词）：It is a relatively complex form of compounding, in which two roots are blended by joining the initial part of the first root and the final part of the second root, or by joining the initial parts of the two roots.5. Abbreviation or clipping:(缩写）A new word is created by cutting the final part or cutting the initial part.6. acronym:(取首字母的缩写词）It is made up from the first letters of the name of an organization, which has a heavily modified (修饰）headword.7. Metanalysis:(再分化）It refers to a process through which a division is made where there were note before.8. Back-formation:(逆构词）It refers to an abnormal(非正常）type of word-formation where a shorter word is derived by deleting(去掉）an imagined affix from a longer form already present in the language.9. Analogical creation:(类比造词）It can account for（说明）the co-existence of two forms, regular and irregular, in the conjugation(结合）of some English verbs.10. Borrowing（借用）：English in its development has managed to widen her vocabulary by borrowing words from other languages.11. Phonological change(音变）：It is related to language variation in the phonological system of language. It includes loss,(省音）addition,（加音）assimilation,（同化）dissimilation.（异化）12. Grammatical change: Changes in both morphology（词法）and syntax（句法）are listed under this heading.13. Semantic change:(语义变化）It includes broadening,（语义扩大）narrowing,（语义缩小）meaning shift,（意义转化）class shift（词性转换）and folk etymology.（词源变化）14. Orthographic change ：(正字法）Changes can also be found at the graphetic level.分享给。
Replace thisfile with prentcsmacro.sty for your meeting,or with entcsmacro.sty for your meeting.Both can befound at the ENTCS Macro Home Page.Hybrid Logics and Ontology Languages Ian Horrocks1Birte Glimm2Ulrike Sattler3School of Computer ScienceUniversity of ManchesterManchester,UKAbstractDescription Logics(DLs)are a family of logic based knowledge representation formalisms.Although they have a range of applications,they are perhaps best known as the basis for widely used ontology languages such as OIL,DAML+OIL and OWL,the last of which is now a World Wide Web Consortium(W3C) recommendation.SHOIN,the DL underlying OWL DL(the most widely used“species”of OWL),includes familiar features from hybrid logic.In particular,in order to support extensionally defined classes,SHOIN includes nominals:classes whose extension is a singleton set.This is an important feature for a logic that is designed for use in ontology language applications,because extensionally defined classes are very common in ontologies.Binders and state variables are another feature from Hybrid Logic that would clearly be useful in an ontology language,but it is well known that adding this feature to even a relatively weak language would lead to undecidability.However,recent work has shown that this feature could play a very useful role in query answering,where the syntactic structure of queries means that the occurrence of state variables is restricted in a way that allows for decidable reasoning.Keywords:Description Logic,Hybrid Logic,Tableaux Reasoning1IntroductionDescription Logics(DLs)are a family of logic based knowledge representation for-malisms.Although they have a range of applications(e.g.,configuration[23],and information integration[7]),they are perhaps best known as the basis for widely used ontology languages such as OIL,DAML+OIL and OWL[14],the last of which is now a World Wide Web Consortium(W3C)recommendation[4].The OWL specification describes three language“species”,OWL Lite,OWL DL and OWL Full,two of which(OWL Lite and OWL DL)are based on expressive description logics.4The decision to base these languages on DLs was motivated by a requirement that key inference problems(such as ontology consistency)be decidable,and hence that it should be possible to provide reasoning services to support ontology design and deployment[14].1Email:horrocks@2Email:glimm@3Email:sattler@4OWL Full uses the same language vocabulary as OWL DL,but does not restrict its use to“well formed formulae”.c 2006Published by Elsevier Science B.V.OWL Lite and OWL DL are based on the DLs SHIF and SHOIN respectively—in fact,OWL Lite is just a syntactic subset of OWL DL[14].5 SHOIN,the DL underlying OWL DL,includes familiar features from hybrid logic. In particular,in order to support extensionally defined classes,SHOIN includes nominals:classes whose extension is a singleton set.This is an important fea-ture for a logic that is designed for use in ontology language applications,because extensionally defined classes are very common in ontologies.For example,OWL provides a oneOf class constructor that allows users to define classes by enumerat-ing their members,e.g.,when describing a class such as EUCountries by enumerating its members,i.e.,{Austria,...,UnitedKingdom}(such an enumeration is equivalent to a disjunction of nominals).This allows applications to infer,e.g.,that persons who only visit EUCountries can visit at most25countries.Singleton classes are also widely used in ontologies,e.g.,in the well known OWL Wine ontology(see /TR/owl-guide/wine.rdf),where they are used for colours, grape types,wine producing regions,etc.Binders(in particular the↓binder)and state variables are another feature from Hybrid Logic that would clearly be useful in an ontology language,but it is well known that adding this feature to even a relatively weak language would lead to undecidability[5].Recent work has shown,however,that this feature could play a very useful role in query answering,where the syntactic structure of queries means that the occurrence of state variables is restricted in a way that allows for decidable reasoning.2PreliminariesThe correspondence between modal and description logics has long been understood [29].The basic propositionally closed DL ALC is equivalent to the propositional modal logic K(m).The logic S extends ALC with transitive roles,and so can be thought of as the union of K(m)and K4(m).SHOIQ extends S with a hierarchy of roles(H),i.e.,the ability to assert implications w.r.t.modalities,nominals(O), inverse roles(I),i.e.,converse modalities,and qualified cardinality constraints(Q), i.e.,graded modalities.The DL SHOIN is a restricted version of SHOIQ,which is obtained by allowing only unqualified cardinality constraints,i.e.,graded modalities can range over the concept only.We will now briefly introduce the syntax,semantics,and inference problems of the DL SHOIQ;full details can be found in[16].Definition2.1Let R be a countable set of role names with both transitive and normal role names R+∪R P=R,where R P∩R+=∅.The set of SHOIQ-roles(or roles for short)is R∪{R−|R∈R}.A role inclusion axiom is of the form R S, for two roles R and S.A role hierarchy is afinite set of role inclusion axioms.An interpretation I=(∆I,·I)consists of a non-empty set∆I,the domain of I,and a function·I which maps every role to a subset of∆I×∆I such that,for 5OWL also includes datatypes,a simple form of concrete domain[2].These can,however,be treated exactly as in SHOQ(D)/SHOQ(D n)[15,11,27],so we will not complicate matters by considering them here.P∈R and R∈R+,x,y ∈P I iff y,x ∈P−I,and if x,y ∈R I and y,z ∈R I,then x,z ∈R I.An interpretation I satisfies a role hierarchy R if R I⊆S I for each R S∈R; such an interpretation is called a model of R.Definition2.2Let N C be a countable set of concept names with a subset N I⊆N C of nominals.The set of SHOIQ-concepts(or concepts for short)is the smallest set such that1.every concept name C∈N C is a concept,2.if C and D are concepts and R is a role,then(C D),(C D),(¬C),(∀R.C), and(∃R.C)are also concepts(the last two are called universal and existential restrictions,resp.),and3.if C is a concept,R is a simple role6and n∈I N,then( nR.C)and( nR.C) are also concepts(called atmost and atleast number restrictions).The interpretation function·I of an interpretation I=(∆I,·I)maps,additionally, every concept to a subset of∆I such that(C D)I=C I∩D I,(C D)I=C I∪D I,¬C I=∆I\C I, o I=1for all o∈N I,(∃R.C)I={x∈∆I|R I(x,C)=∅},(∀R.C)I={x∈∆I|R I(x,¬C)=∅},( nR.C)I={x∈∆I| R I(x,C) n},and( nR.C)I={x∈∆I| R I(x,C) n},where M is the cardinality of a set M and R I(x,C)is defined as{y| x,y ∈R I and y∈C I}.We sometimes use(=nR.C)as an abbreviation for( nR.C) ( nR.C).For C and D(possibly complex)concepts,C˙ D is called a general concept inclusion(GCI),and afinite set of GCIs is called a TBox.Let N A be a countable set of individual names.For a,b∈N A,C a(possibly complex)concept,and R a role,an individual assertion is of the form C(a)or R(a,b),and afinite set of individual assertions is called an ABox.For T a TBox and A and ABox,a pair T,A is called a Knowledge Base(KB).Individual names can also be seen as a weak form of nominals,and it is well known that,in the presence of nominals,an ABox can be expressed in terms of TBox axioms[28].For a DL with nominals, we can therefore assume that each individual assertion C(a)is an abbreviation for a˙ C with a∈N I and each assertion R(a,b)is an abbreviation for a˙ ∃R.b with a,b∈N I.Given a nominal o with o I={x}for some individual x∈∆I,we will 6A simple role is a role that is neither transitive nor has a transitive sub-role;restricting number restrictions to simple roles is required in order to yield a decidable logic[17].sometimes abuse our notation by treating o as an individual name such that o I=x. Please note that we do not make a unique name assumption:two nominals(resp. individual names)might have the same interpretation.An interpretation I satisfies a GCI C˙ D if C I⊆D I,it satisfies an individual assertion C(a)if a I∈C I,and it satisfies an individual assertion R(a,b)if a I,b I ∈R I.An interpretation I satisfies a TBox T(resp.ABox A)if I satisfies each GCI in T(resp.each individual assertion in A),and I satisfies a KB K= T,A if I satisfies both T and A;such an interpretation is called a model of T(resp.A,K).A concept C is satisfiable w.r.t.a role hierarchy R and a KB K if there is a model I of R and K withC I=∅.Such an interpretation is called a model of C w.r.t.R and K.A conceptD subsumes a concept C w.r.t.R and K(written C R,K D)if C I⊆D I holds in every model I of R and K.Two concepts C,D are equivalent w.r.t.R and K(written C≡R,K D)if they are mutually subsuming w.r.t.R and K.For an individual name a,we say that a is an instance of a concept C w.r.t.a role hierarchy R and a KB K if a I∈C I for every model I of R and K,and,for a,b∈I N A,the pair(a,b)is an instance of a role R if a I,b I ∈R I for every model I of R and K.3Reasoning with NominalsA key motivation for basing OWL DL on Description Logics was in order to ex-ploit both theoretical results and implemented DL reasoning systems.Interestingly, though,at the time of the standardisation of OWL there was no known“practical”algorithm for deciding the satisfiability of a SHOIQ KB(where by practical,we mean a goal directed procedure that is likely to behave well on typical ontology de-rived problems[33,15]).As a consequence,no implemented systems were available either.Although,via correspondences with other logics,SHOIQ was known to be decidable[33,26],it proved difficult to extend existing tableaux decision procedures to deal with SHOIQ.The problem is not caused by nominals alone—decision pro-cedures for SHOQ(i.e.,SHOIQ minus inverse roles)and SHOI(i.e.,SHOIQ minus number restrictions)were already known—but by the combination of nomi-nals with inverse roles and number restrictions.In order to see why this is problematical,it is useful tofirst consider some of the features of SHIQ(i.e.,SHOIQ minus nominals),and of the tableaux decision procedure for SHIQ.One reason why DLs(and propositional modal and dynamic logics)enjoy good computational properties,such as being robustly decidable,is that they have some form of tree model property[35,10],i.e.,if an ontology is consistent,then it has a model with a tree-like relational structure.This feature is crucial in the design of tableaux algorithms,allowing them to search only for tree-like models.More precisely,DL tableaux algorithms decide consistency of an ontology by trying to construct an abstraction of a model for it,a so-called“completion graph”. In a completion graph,each node x represents one or more individuals,and is la-belled with a set of concepts which the individuals represented by x are instances of;each edge x,y represents one or more pairs of individuals,and is labelled witha set of roles which the pairs of individuals represented by x,y are instances of. The algorithm works by initialising the graph with one node for each individual name/nominal in the input KB,and using a set of expansion rules to syntactically decompose concepts in node labels;each such rule application can add new con-cepts to node labels and/or new nodes and edges to the completion graph,thereby explicating the structure of a model.The rules are repeatedly applied until either the graph is fully expanded(no more rules are applicable),in which case the graph can be used to construct a model that is a witness to the satisfiability of the input KB,or an obvious contradiction(called a clash)is discovered(e.g.,both C and¬C in a node label),proving that the completion graph does not correspond to a model. The input KB is consistent iffthe rules(some of which are nondeterministic)can be applied in such a way as to produce a fully expanded and clash free completion graph.For logics with the tree model property,we can restrict our search/construction to tree-shaped completion graphs.For expressive DLs,this restriction is crucial, since tableaux algorithms for them employ a cycle detection technique called block-ing in order to guarantee termination.This is of special interest for SHIQ,where the interaction between inverse roles and number restrictions results in the loss of thefinite model property,i.e.,there are consistent ontologies that only admit infinite models.On such an input,the SHIQ tableaux algorithm generates afinite,tree-shaped completion graph that can be unravelled into an infinite tree model,and where a node in the completion graph may stand for infinitely many elements of the model.Even when the language includes nominals,but excludes one of number re-strictions or inverse roles[15,13],or if only individual names instead of nominals are allowed[18],we can work on forest-shaped completion graphs,with each nominal (individual)being the root of a tree-like section;this causes no inherent difficulty as the size of the non-tree part of the graph is restricted by the number of individual names/nominals in the input.The difficulty in extending the SHOQ or SHIQ algorithms to SHOIQ is due to the interaction between nominals,number restrictions,and inverse roles,which leads to the almost complete loss of the tree model property,and causes the complexity of the ontology consistency problem to jump from ExpTime to NExpTime[32].To see this interaction,consider an ontology containing the following two axioms involving a nominal o and a non-negative integer n:F˙ ∃U−.oo˙ ( n U.F)Thefirst statement requires that,in a model of this ontology,every instance of F has an incoming U-edge from o,while the second statement restricts the number of U-edges going from o to instances of F to at most n.The nominal o thus acts as a so-called“spy point”for instances of the concept F(o can“see”every instance of F),and the number restriction on the inverse of U imposes an upper bound of n on the number of instances of F.If we add further axioms,we might need to consider arbitrarily complex relational structures amongst instances of F.For example,ifwe add the following axiom,then there must be exactly n instances of F,and each instance of F is necessarily R-related to every instance of F(including itself):F˙ (=nR.F).Similarly,the following axiom would enforce S-cycles over instances of F:F˙ (=1S.F) (=1S−.F).Hence a tableaux algorithm for SHOIQ needs to be able to handle arbitrarily com-plex relational structures,and thus we cannot restrict our attention to completion trees or forests.To complicate matters even more,recall that there are even SHIQ axioms that enforce the existence of an infinite number of instances of a concept.For example, the concept¬N ∃T.N is satisfiable w.r.t.the following axiom,but only in models with infinitely many instances of N:N˙ ( 1T−.(A ¬A)) ∃T.N.Now consider an ontology that contains,amongst others,all the above mentioned axioms.The consistency of this ontology then crucially depends on the relations enforced between instances of F and N.For example,the additional axiomsN˙ ∃V.F andF˙ ( kV−.N)yield an inconsistent ontology since our at most n instances of F cannot play the rˆo le of V-fillers for infinitely many instances of F when each of them can be the V-filler of at most k instances of N.Summing up,a tableaux algorithm for SHOIQ needs to be able to handle both arbitrarily complex relational structures andfinite tree structures representing in-finite trees,and to make sure that all constraints are satisfied—especially number restrictions on relations between these two parts—while still guaranteeing termina-tion.Two key intuitions have allowed us to devise a tableaux algorithm that meets all of these requirements.Thefirst intuition is that,when extending a SHOIQ completion graph,we can distinguish those nodes that may be arbitrarily intercon-nected(so-called nominal nodes)from those nodes that still form a tree structure (so-called blockable nodes).Moreover,restrictions on the ways in which nominal nodes can be created allow us tofix a(double exponential)upper bound on the number of nominal nodes that can occur in a completion graph.This allows us to restrict blocking,and hence unravelling,to blockable nodes,i.e.,blockable nodes may represent an infinite number of individuals,whereas a nominal node represents exactly one individual in the model.This allows us tofix an upper bound on the size of a completion graph,but it is still not enough to guarantee termination,as we may repeatedly create and merge nodes.For example,given the axiom:C (∃R.C) (∀R−.o),where o is a nominal,constructing a completion graph for C could lead to a se-quence of three nodes,say x0,x1and x2,with{r}labelled edges connecting x0with x1and x1with x2,and with the nodes labelled{C,(∃R.C),(∀R−.o),o}, {C,(∃R.C),(∀R−.o),o}and{C,(∃R.C),(∀R−.o)}respectively.The tableaux rule dealing with nominals might then be applied in order to merge x1into x0(reflecting the semantics of nominals),but applying the rule dealing with existential restric-tions to(∃R.C)in the label of x2would lead to the creation of a new{R}labelled edge connecting x2to a new node,say x3,which,after some additional expansion, would have the same label as x2.Applying the rule dealing with value restrictions to(∀R−.o)in the label of x3would cause o to be added to the label of x2,and allow the above process to be repeated w.r.t.x0,x2and x3.This(particular version of the)problem was identified by Baader,and called,for obvious reasons,a“yo-yo”[2,3].This problem is normally addressed by discarding sub-trees below a node when-ever it is merged with another node.In the above case,for example,this would result in x2(and the edge from x1to x2)being removed from the graph when x1is merged with x0;the resulting graph is fully expanded,and so the algorithm would terminate.This only works,however,in tree-shaped parts of the completion graph, as the notion of a sub-tree is otherwise not well defined.This can lead to problems if nominal nodes are created and merged.The second and crucial intuition is that this problem can be overcome by“guess-ing”the exact number of new nominal nodes created as the result of interactions between existing nominal nodes,inverse roles,and number restrictions.This guess-ing is implemented by a new expansion rule,the NN-rule.When applied to a relevant( nR.C)concept in the label of a nominal node x,the NN-rule chooses (non-deterministically)an integer m,such that1 m n,adds( mR.C)to the label of x,and generates m new nominal nodes,all of which are pairwise disjoint, and m new{R}labelled edges leading from x to the new nominal nodes.As they are pairwise disjoint,none of these nodes can be merged,and adding( mR.C) means that no more nominal nodes can be added as a result of concepts of the form ( nR.C)in the label of x.Termination is now guaranteed by the upper bound on the number of nominal nodes and the use of standard blocking techniques for the blockable nodes.The non-determinism introduced by this rule will clearly be prob-lematical for large values of n,but large values in number restrictions are already known to be problematical for SHIQ.Moreover,the rule has excellent“pay as you go”characteristics:in case number restrictions are functional(i.e.,where n is 1),7the new rule becomes deterministic;in case there are no interactions between number restrictions,inverse roles,and nominals,the rule will never be applied;in case there are no nominals,the new algorithm behaves like the algorithm for SHIQ; and in case there are no inverse roles the new algorithm behaves like the algorithm 7The DL obtained from SHIQ by allowing only functional number restrictions is called SHIF.There are many realistic SHIF ontologies;see, e.g.,the DAML ontology library at /ontologies/for SHOQ.Indeed,recent implementations of this algorithm in Fact++and Pellet[34,30] show promising behaviour:for example,the Wine ontology8can now be classified, and both reasoners only take a few seconds for its classification,despite the fact that it is in SHOIF and contains206nominals.Recent work has also shown how a similar technique can be used in order to extend resolution based reasoning procedures to deal with SHOIQ[21].4Answering Conjunctive QueriesExisting DL reasoners(for example FaCT++[34],KAON2[20],Pellet[30]and Racer Pro[12])provide automated reasoning support for checking concepts for sat-isfiability and subsumption,and also for answering queries that retrieve instances of concepts and roles.In many applications,however,a more powerful query language is required.An obvious approach would be to extend Datalog style conjunctive queries to DLs,and this has led to studies of the problem of conjunctive query answering for DLs[22,19].As usual,we will restrict our attention to the problem of answering Boolean conjunctive queries,i.e.,queries where the answer is either true or false;it is well known how to reduce non-Boolean queries(i.e.,queries where the answer is a set of tuples of individuals)to Boolean ones.Definition4.1A Boolean conjunctive query q has the form ←conj1( y; c)∧...∧conj n( y; c),where y is a vector of variables and c is a vector of individual names.We call T(q)= y∪ c the set of terms in q,9and we call each conj i( y; c)for1≤i≤n an atom.Atoms are either concept or role atoms:a concept atom has the form C(t), and a role atom the form R(t,t ),for{t,t }⊆T(q),C a possibly complex concept, and R a role.The semantics of Boolean conjunctive queries is defined in terms of assignments. Let K be an knowledge base,I=(∆I,·I)a model for K,and q a Boolean conjunctive query.An assignment in I is a mapping·A:T(q)→∆I.We say that q is true in I and write I|=q if there exists an assignment·A in I s.t.t A∈{t I}for every individual name t∈ c,t A∈C I for every concept atom C(t)in q,and t A,t A ∈R I for every role atom R(t,t )in q.We call such an assignment a q-mapping w.r.t.I. If I|=K implies I|=q for all models I of K,then we say that q is true in K,and write K|=q;otherwise we say that q is false in K,and write K|=q.The development of a decision procedure for conjunctive query answering in expressive DLs is still a partially open problem.Grounded conjunctive queries for SHIQ and/or SHOIQ are supported by KAON2,Pellet,and Racer’s query language nRQL[24,12].However,the semantics of grounded queries is different from the usually assumed open-world semantics in DLs since existentially quantified variables are always bound to named individuals.Various decision procedures[9,25,19]are known for restricted fragments of con-junctive queries.The most common restriction used in order to obtain a decision8Available at /2001/sw/WebOnt/guide-src/wine.rdf.9For readability,we sometimes abuse our notation and refer to y as a set.When referring to a vector y as a set,we mean the set{y i|y i occurs in y}.procedure is that each role that occurs in the query must be a simple role,i.e.,a role that is neither transitive nor has a transitive sub-role.A commonly used technique is to reduce the problem of answering a Boolean conjunctive query to the problem of deciding concept satisfiability w.r.t.a KB.The idea is to view the query as a graph and,starting from leaf nodes,to“roll up”the graph into a single atom of the form C(t)such that the query is true in a KB K iff¬C is unsatisfiable w.r.t.K[9,6,19,31].For example,the queryq= ←R(t,t )∧S(t ,t )∧C(t )could be rolled up into←(∃R.(∃S.C))(t)such that K|=q iff¬(∃R.(∃S.C))is unsatisfiable w.r.t.K.Unfortunately,the above described technique is only able to deal with tree shaped queries,i.e.,those where the query graph does not contain a cycle,because the above mentioned tree model property means that DL concepts cannot capture cyclic relationships.Even nominals cannot express the arbitrary cyclic structures that can occur in a query[31].Recent work has,however,shown how an extension of DLs with a restricted form of the↓binder operator and state variables can enable an extension of the rolling-up technique to deal also with cyclic queries[9].It is easy to see how these features can be used to roll up an arbitrary query into a suitable concept.For example,the queryq= ←R(t,t )∧S(t ,t)could be rolled up into the concept←(↓t.(∃R.(∃S.t)))(t)Although the↓binder already makes the DL ALC undecidable,when used in query answering the↓binder occurs only in a very restricted form,and the resulting extension of SHOIQ is still decidable[8]if only simple roles are used in the query. Moreover,it is relatively easy to extend a tableaux algorithm for SHOIQ to one with this restricted form of the↓binder.However,although it is now known that conjunctive query answering for SHIQ(without any restrictions)is decidable[1], it is not clear how to extend the↓binder technique to queries in which non-simple roles occur in a cycle.In this case,the binder interacts with blocking in a way that makes termination problematical.5DiscussionAs we have seen,description logics underlying state of the art ontology languages include nominals,a well known feature of hybrid logic.Nominals are an important feature of ontology languages,as extensionally defined and singleton classes are common in ontologies.Recent advances in DL reasoning techniques have shown how the tableaux algorithm for SHIQ,which is widely used in DL reasoners,can be extended to deal with nominals,and implementations of the shown algorithmhave already exhibited promising results.Moreover,it has also been shown that a similar technique can be applied to resolution based reasoning procedures.The↓binder,another familiar feature of hybrid logics,has also been investi-gated in the context of DL reasoning,in this case algorithms for conjunctive query answering.It has been shown that an extension of SHOIQ to include a restricted form of this feature is not only decidable,but is extremely useful in the context of conjunctive query answering.Currently,the decision procedure is restricted to con-junctive queries with only simple roles and extending the technique to unrestricted conjunctive queries is still an open problem.References[1]Anonymous.Conjunctive query answering for SHIQ.Technical report,2006.http://ijcai.tripod.com/SHIQ-CQ.pdf.[2]F.Baader and P.Hanschke.A schema for integrating concrete domains into concept languages.InProc.of the12th Int.Joint Conf.on Artificial Intelligence(IJCAI’91),pages452–457,1991.[3]F.Baader and U.Sattler.An overview of tableau algorithms for description logics.Studia Logica,69(1):5–40,Oct.2001.[4]S.Bechhofer,F.van Harmelen,J.Hendler,I.Horrocks,D.L.McGuinness,P.F.Patel-Schneider,and L.A.Stein.OWL web ontology language reference.W3C Recommendation,10February2004.Available at /TR/owl-ref/.[5]P.Blackburn and J.Seligman.Hybrid languages.Journal of Logic,Language and Information,4(3):251–272,1995.Special issue on decompositions offirst-order logic.[6]D.Calvanese,G.De Giacomo,and M.Lenzerini.On the decidability of query containment underconstraints.In Proc.of the17th ACM SIGACT SIGMOD SIGART Symp.on Principles of Database Systems(PODS’98),pages149–158,1998.[7]D.Calvanese,G.De Giacomo,M.Lenzerini,D.Nardi,and R.Rosati.Description logic framework forinformation integration.In Proc.of the6th Int.Conf.on Principles of Knowledge Representation and Reasoning(KR’98),pages2–13,1998.[8]B.Glimm,I.Horrocks,and U.Sattler.Conjunctive query answering for description logics with transitiveroles.In Proc.of the2006Description Logic Workshop(DL2006),pages3–14.CEUR Electronic Workshop Proceedings,/Vol-189/,2006.[9]B.Glimm,I.Horrocks,and U.Sattler.Conjunctive query answering for the description logicSHOIQ.Technical report,The University of Manchester,2006./∼horrocks/Publications/download/2006/GlHS06b.pdf.[10]E.Gr¨a del.The decidability of guardedfixed point logic.In J.Gerbrandy,M.Marx,M.de Rijke,andY.Venema,editors,Essays Dedicated to Johan van Benthem on the Occasion of his50th Birthday.Amsterdam University Press,1999.[11]V.Haarslev and R.M¨o ller.Description of the RACER system and its applications.In Proc.of the2001Description Logic Workshop(DL2001),pages132–141.CEUR Electronic Workshop Proceedings, /Vol-49/,2001.[12]V.Haarslev,R.M¨o ller,and M.Wessel.Querying the Semantic Web with Racer+nRQL.In Proc.ofthe KI-2004Intl.Workshop on Applications of Description Logics(ADL’04),2004.[13]J.Hladik and J.Model.Tableau systems for SHIO and SHIQ.In Proc.of the2004Description LogicWorkshop(DL2004).CEUR,2004.Available from .[14]I.Horrocks,P.F.Patel-Schneider,and F.van Harmelen.From SHIQ and RDF to OWL:The makingof a web ontology language.J.of Web Semantics,1(1):7–26,2003.[15]I.Horrocks and U.Sattler.Ontology reasoning in the SHOQ(D)description logic.In Proc.of the17th Int.Joint Conf.on Artificial Intelligence(IJCAI2001),pages199–204.Morgan Kaufmann,Los Altos,2001.[16]I.Horrocks and U.Sattler.A tableaux decision procedure for SHOIQ.In Proc.of the19th Int.JointConf.on Artificial Intelligence(IJCAI2005),pages448–453,2005.。
