4 Pages

lx502f06-1a-intro

Course: LX 502, Fall 2009
School: BU
Rating:
 
 
 
 
 

Word Count: 1307

Document Preview

dog Your ate my homework. CAS LX 502 1a. Introduction What does this sentence mean? Is it true? What is the current status of my homework? What was the prior status of my homework? Your dog will have eaten my homework. Your dog must have eaten my homework. Your dog might have eaten my homework. Levels of linguistic knowledge Native speakers have a complex system of knowledge. Phonology Morphology...

Register Now

Unformatted Document Excerpt

Coursehero >> Massachusetts >> BU >> LX 502

Course Hero has millions of student submitted documents similar to the one
below including study guides, practice problems, reference materials, practice exams, textbook help and tutor support.

Course Hero has millions of student submitted documents similar to the one below including study guides, practice problems, reference materials, practice exams, textbook help and tutor support.
dog Your ate my homework. CAS LX 502 1a. Introduction What does this sentence mean? Is it true? What is the current status of my homework? What was the prior status of my homework? Your dog will have eaten my homework. Your dog must have eaten my homework. Your dog might have eaten my homework. Levels of linguistic knowledge Native speakers have a complex system of knowledge. Phonology Morphology Syntax Semantics Some data our theory should account for Linguistic expressions have actual meanings The English sentence Camels have humps means that camels have humps. The French sentence Les chameaux ont des bosses means that camels have humps. Our goal is to understand and characterize this complex system. The English sentence Camels have humps does not mean that reptiles have wings. Some data our theory should account for Linguistic expressions can be ambiguous. The English sentence Pedro jumped from the top of the bank has two meanings. Some data our theory should account for Some linguistic expressions are felt to be anomalous. Colorless green ideas sleep furiously. My toothbrush is pregnant. Hammers frighten rocks. The English sentence Mad dogs and Englishmen go out in the noonday sun has two meanings. The English sentence John saw her duck has two meanings. 1 Some data our theory should account for These are true in the same situations (synonymous): The wug is under the pim. The pim is above the wug. Some data our theory should account for Boris ate an apple. Boris ate. Boris is too clever to catch Rocky. Boris is too clever to catch. And not in this one (contradiction): The wug is not under the pim. At least two distinct kinds of situations make this true: Every wug is under three pims. And this must also be true of the same kind of situation: Every wug is under two pims. Productivity We can know what a sentence means even if we havent heard that sentence before. Productivity We can create novel meaningful sentences. S1 and S2 is a sentence where S1 and S2 are If we knew what a wug and a pim were, we would know what the sentence The pim is under the wug would mean. There must be an algorithm for figuring out what a sentence means (like the algorithm for adding two novel numbers). sentences (and S1 and S2 is true when both S1 and S2 are true). Sa = Pat left Sb = Tracy cried Sc = Chris ate a sandwich S1 = Sa and Sb = Pat left and Tracy cried S2 = S1 and Sc = Pat left and Tracy cried and Chris ate a sandwich Compositionality In order to be able to do this is must be the case that meaning is compositional. That is: A noble spirit embiggens the smallest man Fax, refax, faxer, refaxer, unrefaxeristic. Unzippable [Un-zip]-able (able to be unzipped) Un-[zip-able] (unable to be zipped) The meaning of the whole is determined by the meaning of the parts and how those parts are arranged. 2 Connecting to the world We use language to talk about things. There is some kind of connection to the world outside of language. Connecting to the world Truth and falsity is something like this as well. For example, we can use names and noun phrases to refer to individuals. We use Pat to refer to Pat the individual; Pat denotes Pat the individual. Whether a sentence is true or false depends on The actual meaning of the sentence How things are in the outside world Levels of description We can talk about language from a number of different perspectives. speaking or writing. Propositions A proposition can be true or false, depending on A proposition is generally a combination of a grilled (Pat, the sandwich) the situation it is being considered with respect to. predicate (a verb or a property) and one or more arguments (referential expressions). one way (Pat grilled the sandwich, The sandwich was grilled by Pat), so are they more abstract than sentences. An utterance is a piece of language, created, e.g., by A sentence is an abstract arrangement of words. Two people can say the same sentence (resulting in two different utterances). A proposition is a complete thought. The sentences Pat ate the sandwich and The sandwich was eaten by Pat express the same proposition. One is true if and only if the other is true. Propositions can often be expressed in more than Language in context Sentences can be said to have a literal and, sometimes, a non-literal meaning. My feet are killing me. Language in context When we speak of semantics, we are generally speaking of meaning abstracted away from speakers and individual utterances. The context-independent meaning. When we speak of pragmatics, we are speaking of language use and language users. The motivations for using a particular sentence. Could you pass the salt? Its getting late. I have two children. Mr. Stevens command of English is excellent, and his attendance at class has been regular. Literally false, but the speaker presumably intended to communicate something, so we must work out what it is. 3 Mental models Prince Charles is British. This is true. Theres a property, to be Mental models We can talk about much more than just the real world as it actually is. If Prince Charles is not British, I will eat my hat. British, that holds of the guy in the world that Prince Charles denotes. Imagine a world much like ours, but in which the guy Prince Charles denotes does not possess the property to be British. In that world, I will also eat my hat (if the entire proposition is true). Last night I dreamt I was Prince Charles. James Bond is British. This is also true. But who is the property to be British true of? Mental models Denotation and properties seem to be evaluated against a model of the world, which maybut need notcorrespond to how the world actually is. We can shift between models easily, to talk about fiction, to talk about (possibly mistaken) beliefs, to talk about wishes, dreams. The most coherent way to think about what names like James Bond denote is to take them to denote individuals in a mental model, where the model might (or might not) be one corresponding to the real world. Object language and ...

Find millions of documents on Course Hero - Study Guides, Lecture Notes, Reference Materials, Practice Exams and more. Course Hero has millions of course specific materials providing students with the best way to expand their education.

Below is a small sample set of documents:

Pittsburgh - CS - 2740
University of Pittsburgh CS 2740 Knowledge Representation Professor Milos HauskrechtHandout 2 September 10, 2008Problem assignment 1Due: Wednesday, September 17, 2008LispThe goal of this assignment is to practice your lisp programming skills
Pittsburgh - CS - 441
CS 441 Discrete Mathematics for CS Lecture 11Sets and set operationsMilos Hauskrecht milos@cs.pitt.edu 5329 Sennott SquareCS 441 Discrete mathematics for CSM. HauskrechtCourse administrationHomework 3: Due today Homework 4: Due next week
Pittsburgh - CS - 2740
CS 2740 Knowledge Representation Lecture 4Propositional logicMilos Hauskrecht milos@cs.pitt.edu 5329 Sennott SquareCS 2740 Knowledge RepresentationM. HauskrechtAdministration Homework assignment 1 is out Due next week on Wednesday, Septem
Pittsburgh - CS - 441
CS 441 Discrete Mathematics for CS Lecture 8Methods of ProofMilos Hauskrecht milos@cs.pitt.edu 5329 Sennott SquareCS 441 Discrete mathematics for CSM. HauskrechtCourse administration Homework 1 and Homework 2: Due today Homework 3 out to
Pittsburgh - CS - 1571
CS 1571 Introduction to AI Lecture 13Propositional logicMilos Hauskrecht milos@cs.pitt.edu 5329 Sennott SquareCS 1571 Intro to AIM. HauskrechtAnnouncements Homework assignment 5 is out Midterm exam: Wednesday, October 25, 2006 Course web p
Pittsburgh - CS - 3750
Probabilistic Principal Component Analysis and the E-M algorithmThe Minh Luong CS 3750 October 23, 2007Outline Probabilistic Principal Component Analysis Latent variable models Probabilistic PCA Formulation of PCA model Maximum likelihood est
Pittsburgh - CS - 441
CS 441 Discrete Mathematics for CS Lecture 19Summations, CardinalityMilos Hauskrecht milos@cs.pitt.edu 5329 Sennott SquareCS 441 Discrete mathematics for CSM. HauskrechtCourse administration Homework 5 is due today Homework 6 is out Due o
Pittsburgh - CS - 2740
University of Pittsburgh CS 2740 Knowledge Representation Professor Milos HauskrechtHandout 5 October 1, 2008Problem assignment 4Due: Wednesday, October 8, 2008UnicationThe unication process in FOL aims to nd the most general substitution tha
Pittsburgh - CS - 2740
CS 2740 Knowledge representation Lecture 1Knowledge representationMilos Hauskrecht milos@cs.pitt.edu 5329 Sennott SquareCS 2740 Knowledge representationM. HauskrechtCourse administrationInstructor: Milos Hauskrecht 5329 Sennott Square milo
Pittsburgh - CS - 3750
CS 3750 Machine Learning Lecture 11Monte Carlo inferenceMilos Hauskrecht milos@cs.pitt.edu 5329 Sennott SquareCS 3750 Advanced Machine LearningMarkov chain Monte Carlo Likelihood weighting: samples are generated according to Q and every sampl
Pittsburgh - CS - 2710
CS 2710 Foundations of AI Lecture 8Adversarial searchMilos Hauskrecht milos@cs.pitt.edu 5329 Sennott SquareCS 2710 Foundations of AIGame search Game-playing programs developed by AI researchers since the beginning of the modern AI era Progr
Pittsburgh - CS - 1571
CS 1571 Introduction to AI Lecture 12Adversarial searchMilos Hauskrecht milos@cs.pitt.edu 5329 Sennott SquareCS 1571 Intro to AIM. HauskrechtAnnouncements Homework assignment 4 is out Programming and experiments Simulated annealing + Gen
Pittsburgh - MAY - 2005
FEATUREAN EDITOR ON A MISSION BY JESSICA MESMANSO YOU WANT TOCHANGETHE WORLD?HSome people might say Ive reached a point in my life where Im not afraid. But Ive never been afraid, says Catherine DeAngelis.PHOTOGRAPHY |ere comes Dr. De, h
Pittsburgh - EXP - 1593
Pittsburgh - EXP - 1652
Internet WavesDisruptive Technologies> Cool ToolsDigging Digsby and Tweeting Locally> Database ReviewSmart Search Engines Find Best Facts Page 26Page 19Page 41iJune 2008Vol. 25 I Issue 6The Newspaper for Users and Producers of Elect
Pittsburgh - SIS - 793
MARLA MISEKMirror Image InternetThrough the Looking GlassIike the heroine of Lewis Carroll's Alice's Adventures in Wotiderland, who finds herself in a fantastical place where things are not always as they seem, Intemet users generally have no
Pittsburgh - EXP - 793
MARLA MISEKMirror Image InternetThrough the Looking GlassIike the heroine of Lewis Carroll's Alice's Adventures in Wotiderland, who finds herself in a fantastical place where things are not always as they seem, Intemet users generally have no
Pittsburgh - SIS - 1821
BUILDING COMMUNITY AS A LIBRARY IN A 3D ENVIRONMENTKelly Czarnecki Technology education librarianPublic Library of Charlotte and Mecklenburg County North CarolinaUSA Received January2008It is predicted in a recent reportby GartnerInc that '80% of a
Pittsburgh - EXP - 1821
BUILDING COMMUNITY AS A LIBRARY IN A 3D ENVIRONMENTKelly Czarnecki Technology education librarianPublic Library of Charlotte and Mecklenburg County North CarolinaUSA Received January2008It is predicted in a recent reportby GartnerInc that '80% of a
Pittsburgh - EXP - 325
CLICK FIRST, ASK QUESTIONS LATER: UNDERSTANDING TEEN ONLINE BEHAVIOURJack Goodman CEO Tutoring AustralasiaExamined are how teenagers engage with technology, particularly the internet; what services, sites and programs they find compelling; and how
Pittsburgh - EXP - 335
Pittsburgh - SIS - 1590
Pittsburgh - EXP - 1590
Pittsburgh - SIS - 1834
Social Networking Mixes the Hip with the Proven"THIS YEAR'S EXPLOSIVE GROWTH IN SOCIAL SOFTWARE REMINDS ME OF SHAKESPEARE'S FAMOUS LINE FROM THE TEMPEST, 'WHAT IS PAST IS PROLOGUE."' When Computers in Librariestakes up K. H UVWE the topic of Hip Hig
Pittsburgh - EXP - 1834
Social Networking Mixes the Hip with the Proven"THIS YEAR'S EXPLOSIVE GROWTH IN SOCIAL SOFTWARE REMINDS ME OF SHAKESPEARE'S FAMOUS LINE FROM THE TEMPEST, 'WHAT IS PAST IS PROLOGUE."' When Computers in Librariestakes up K. H UVWE the topic of Hip Hig
Pittsburgh - EXP - 602
Chapter 5Social VideoVideoblogging & YouTuben the past year, viewing and creating video for the Web skyrocketed to the top of many Internet users' lists. The popularity of YouTube, Google Video, and other sharing sites, coupled with the ease of
Pittsburgh - EXP - 116
3 6 I InformationTodayApril 2007www.lntotocfav.coniwww.infotoday.com/linkupYour Personal Guide 1 0 t:he V\#eb :omeA 'del.icio.us' Way to Use Bookmarksregister With del.icio.usby THOMAS PACKThe Folksonomic ApproachEDITED BYLoRAiNF P
Pittsburgh - EXP - 228
practice & researchblogging for beginnersCan the blogging phenomenon offer anything to people with learning disabilities? Yes, says Denise Stokes, who co-ordinates a pioneering group based in CoventryKeywords comrriLinication information techno
Pittsburgh - EXP - 1654
IOWWW.IWR.CO.UK/INFOTECHSOCIAL COMPUTING WITHOUT TEARSReluctant IT departments are playing it safe with social networking, but that could all be about to change, as David Tebbutt explainsfew months ago the IWR blog carried a post about nast\- goi
Pittsburgh - CIDDEFALL - 02
Media IN and FOR the Language Class Presenter: Zsuzsa HorvthIntroductionSponsored by CIDDE Supported by the Robert Henderson Language Media CenterMedia and Technology in the ClassroomRoom with media service: - pick up key (Instructional Media Se
Pittsburgh - TECHWORKSH - 01
Stacks, Buttons, Fields, and Cards in HyperCardCreating New Stacks Find the icon for "Home" or "HyperCard" and double-click on it.If you are on the first card of the Home stack the upper left portion of your screen probably looks something like t
Pittsburgh - TECHWORKSH - 01
1. Basic Hypercard InformationCommands Which Are Used in Almost All Macintosh Programs Command key: (on either side of the Space Bar)+Q +N +O +W +S +Z +X +C +V +PQuit from a program New file (inside Hypercard: "New card") Open file or program (d
Pittsburgh - SUPER - 7
Administrative Issues in Outbreak InvestigationsOR.How to Optimize Your 15 Minutes of FameM. Joan Mallick, R.N., Ph.D.Part AGetting to Know the Media Before an Emergency StrikesBackgroundA The anthrax "outbreak" of 2001 brought theissue
Wisc Stevens Point - ED - 362
Classroom RulesClassroom rules will be set by the students, but guided by the three teachers around five core values: Body Basics, Honesty, Respect, Responsibility, and Caring. Once the students have created rules that they can agree on, the three t
Pittsburgh - LASA - 97
Better Workers, Better Wives: Vocational Education for Poor Women in Turn-of-the-Century Chile1by Elizabeth Quay Hutchison Colby CollegePrepared for delivery at the 1997 Meeting of the Latin American Studies Association, Continental Plaza Hotel,
BU - SPG - 06
NBER WORKING PAPER SERIESABORTION AND SELECTION Elizabeth Oltmans Ananat Jonathan Gruber Phillip B. Levine Douglas Staiger Working Paper 12150 http:/www.nber.org/papers/w12150 NATIONAL BUREAU OF ECONOMIC RESEARCH 1050 Massachusetts Avenue Cambridge
BU - PDFFALL - 07
R&D and productivity: Estimating production functions when productivity is endogenousUlrich Doraszelski Harvard University and CEPR Jordi Jaumandreu Universidad Carlos III and CEPRApril 2007 Preliminary and incomplete. Comments very welcome.Abst
BU - ELBAZ - 1
BU - ELBAZ - 1
BU - ELBAZ - 1
BU - ELBAZ - 1
BU - ELBAZ - 1
BU - ELBAZ - 1
BU - ELBAZ - 1
BU - ELBAZ - 1
BU - ELBAZ - 1
BU - ELBAZ - 1
BU - ELBAZ - 1
BU - ELBAZ - 1
Pittsburgh - LASA - 98
How U.S. Students Perceive Cuba: A Current Cultural ExperienceAviva Chomsky History Department Salem State College Salem, MA 01970 978-542-6389 achomsky@salem.mass.eduPrepared for delivery at the 1998 meeting of the Latin American Studies Associa
Pittsburgh - CS - 1550
Outline of this class How to submit your assignment? A few examples Install NachOS for second assignmentusemainsimple.c usestring.c usefork.c useexecvp.c~jrmst106/public/1550Synch.pl1
Pittsburgh - CS - 2750
CS 2750 Machine Learning Lecture 13Multi-way classificationMilos Hauskrecht milos@cs.pitt.edu 5329 Sennott SquareCS 2750 Machine LearningAdministrative announcementsHomework 6 due on Wednesday Plan for the upcoming month: Homework 7 due on
Pittsburgh - CS - 2750
CS 2750 Machine Learning Lecture 12Nave Bayes classifier & Evaluation frameworkMilos Hauskrecht milos@cs.pitt.edu 5329 Sennott SquareCS 2750 Machine LearningGenerative approach to classificationIdea: 1. Represent and learn the distribution p
Pittsburgh - CS - 3750
CS 3750 Machine Learning Lecture 20Support vector machinesMilos Hauskrecht milos@cs.pitt.edu 5329 Sennott SquareCS 3750 Advanced Machine LearningLinearly separable classesThere is a hyperplane that separates training instances with no error
Pittsburgh - CS - 2750
CS 2750 Machine Learning Lecture 14Support vector machinesMilos Hauskrecht milos@cs.pitt.edu 5329 Sennott SquareCS 2750 Machine LearningOutlineOutline: Fisher Linear Discriminant Algorithms for linear decision boundary Support vector mach
Pittsburgh - CS - 2750
CS 2750 Machine Learning Lecture 21Learning with hidden variables and missing values.Milos Hauskrecht milos@cs.pitt.edu 5329 Sennott SquareCS 2750 Machine LearningLearning probability distributionBasic learning settings: A set of random vari
Pittsburgh - CS - 2710
CS 2710 Foundations of AI Lecture 25Learning probability distributionsMilos Hauskrecht milos@cs.pitt.edu 5329 Sennott SquareCS 2710 Foundations of AIDensity estimationData: D = {D1 , D2 ,., Dn } Di = x i a vector of attribute values Attribut
Pittsburgh - CS - 3710
CS 3710 Advanced Topics in AI Lecture 6Undirected graphical models and factorsMilos Hauskrecht milos@cs.pitt.edu 5329 Sennott SquareCS 3710 Probabilistic graphical modelsFactors Factor: is a function that maps value assignments for a subset o
Pittsburgh - CS - 3710
CS 3710 Advanced Topics in AI Lecture 15Constructing Free Energy Approximations and GBP AlgorithmsBranislav Kveton bkveton@cs.pitt.edu 5802 Sennott SquareCS 3710 Probabilistic graphical modelsContent Why? Belief propagation (BP) Factor
Pittsburgh - CS - 2620
INTERDISCIPLINARYMODELINGANDVISUALIZATIONIntroductionLiz MaraiJanuary 5, 2009Introduction 1/17INTERDISCIPLINARYMODELINGANDVISUALIZATIONVisualization In general: (graphic) any technique for creating images, diagrams, or animat
Pittsburgh - CS - 2620
=Name: James ShoemakerDept: PsychiatryYear (Fr, So, Jr, Sr, Gr<n>): GRE-mail:calceusATyahoo.comPhone(s): Want to learn: I am interested in learning what is possible using current technology, and how that technology can be expanded.Other