What is Intelligence? - natobilc.org

What is Intelligence? - natobilc.org

Norwegian Defence Intelligence and Security School Intelligence Section NATO RESTRICTED Marit Lobben Language adviser Section for Language Contingency Norwegian Defence Academy E-mail: [email protected] Side 1 NATO RESTRICTED Norwegian Defence Intelligence and Security School Intelligence Section Language Aptitude what is it and how do we measure it? The objective of this presentation is not to give you all best solutions to aptitude testing, but: o o o o o To ask some very basic questions: what is language aptitude and can it be measured? To find out what is the State of the Art for language aptitude testing today Why do we need to have good language aptitude tests? What abilities should a LAT measure and why?

To present a portion of a LAT in the piloting stage which is to be used in Norway to select language students for the Norwegian Defence Side 2 NATO RESTRICTED Norwegian Defence Intelligence and Security School Intelligence Section Very little scientific work published mostly old or outdated from 60s and 70s Useful books on language aptitude o o o 1981: Individual Differences & Universals in Language Learning Aptitude. Ed. by Karl C. Diller. Newbury House Publishers, Inc. Rowley, London, Tokyo. 1990: Language Aptitude Reconsidered. Ed. by Thomas S. Parry and Charles W. Stansfield. In series: Language in Education Theory & Practice, vol 74. ERIC/CAL (Center for Applied Linguistics). Prentice Hall Regents, Englewood Cliffs, new Jersey. 2004 : The Neurobiology of Learning Perspectives from Second Language Acquisition. By John H Schuhmann (UCLA), Sheila E. Crowell (Univ. of Washington, Seattle), Nancy E. Jones (UCLA), Namhee Lee (UC at Riverside), Sara Ann Schuchert (MidWilshire Christian Schools, LA), Lee Alexandra Wood (STARR Litigation Services, Iowa). Published by Lawrence Erlbaum Associates, Mahwah, New Jersey, London. Side 3

NATO RESTRICTED Foreign language learning, names and settings Norwegian Defence Intelligence and Security School Intelligence Section Two types of language learning o o Many names: foreign language, second (third, fourth, etc.) language, cf. L2 language learning in the Applied linguistics paradigm All are opposed to first language acquisition (alternative names: native language, mother tongue) Definition of settings for learning a foreign language o o Formal instruction environment (classroom instruction, self-study, intensive language programs) Being in a foreign language environment little is known about learning in this case Side 4 NATO RESTRICTED Language aptitude - what is it? Norwegian Defence Intelligence and Security School Intelligence Section

The concept of aptitude o o o o o o Aptitude: the capability of an individual to learn (enduring characteristics of the individual). Does not include motivation or interest. Assessed before learning is started Achievement: the notion that an individual may have acquired specific capabilities of actual performance. Assessed after the individual has been exposed to a learning program for a certain time For the sake of argument: is there any real distinction between aptitude and achievement? What do we measure: can measures of aptitude be distinguished from measures of achievement? Claim: what we measure when we measure aptitude is really achievement. However, these are logically distinguishable, although aptitude in some degree is dependent on achievement in that such test to some extent is dependent on past learning . Difference: prior to a learning program there should be no significant difference between an aptitude test and an achievement test measuring the outcome. After the learning program, however, there should be significant differences. Side 5 NATO RESTRICTED Norwegian Defence Intelligence and Security School

Intelligence Section Language aptitude, contd. The concept of foreign language aptitude o o Do some individuals have the knack for learning languages? Needs to be kept apart from other factors such as o Motivation, attitude and perseverance Opportunity to learn General verbal intelligence Quality and type of instruction Statistical considerations support the notion that the predictive powers of certain LATs are substantial and a highly important factor in foreign language learning, in addition to other relevant factors (applies to the MLAT and PLAB tests). Side 6 Norwegian Defence Intelligence and Security School Intelligence Section NATO RESTRICTED Do humans have a separate language ability/ language faculty and what is the nature of this

ability? o o A general and unitary ability? A composite collection of a number of independent, specialized abilities? Side 7 NATO RESTRICTED The language faculty Norwegian Defence Intelligence and Security School Intelligence Section Sources of evidence o Linguistic theories Language ability is innate and modular (Chomskian paradigm). Cf. Colourless, green ideas sleep furiously. Cognitive Grammar (Ronald Langacker, Adele Goldberg) and psycholinguistics (psychologist Jean Berko): Language ability is part of a general cognitive ability. Example: The way in which categories are structured in languages (grammatical categories) resemble those the human brain uses for categories in general. Linguistic meaning is encyclopaedic, not restricted to a limited inventory of semantic features:

Usage based language structures. interpreting lexical and grammatical morphemes in the process of historical change 1) frequency of utterance 2) the number of lexical items a linguistic rule applies to is going to directly affect the linguistic structure of a language. Language faculty is not abstract and unchangeable, but closely related to general abilities of memory and learning. Cogntive psychology studies of attention, perception, etc. Side 8 NATO RESTRICTED The language faculty, contd. o Neutrobiological research Norwegian Defence Intelligence and Security School Intelligence Section Aphasia. When a person has had a part of his/her brain damaged, and simultaneously suffers loss of one or more language faculties, one can infer

from this where in the brain certain that particular language ability is localized. Division of labour between parts of the brain. o Brocas area: Expressive (motor) abilities. Wernickes area: Receptive (sensory) region of communicative language, comprehension of speech. Feature detection of language elements. Two principal types of brain cells in the human cortex: pyramidal and star-shaped (local circuit neurons). Continued learning and memory from after puberty and into adult life. Who is right? Studies in aphasia suggest that at least some areas of the brain can be associated with the language faculty. However, this may not preclude the fact that information processing interacting with other parts of the brain is necessary. In order for the modular Chomskian approach to be confirmed, one would think that linguistic sub-disciplines would have to be attributed to physical areas of the brain. This is yet to be seen. Side 9 NATO RESTRICTED Norwegian Defence Intelligence and Security School Intelligence Section

What abilities are relevant to learning languages? o o o Perception Memory Concept formation o Concept formation is a general cognitive ability. Language is different from general conceptual knowledge in that every language consists of a set of conventions of arbitrary sound-meaning combinations. Concepts relevant to language structures The principles of how languages combine meaning units to larger units are finite and can be described. Relevant field: linguistic typology. Certain linguistic features co-vary. 1) word order (SVO, SOV, VSO) will predict whether a) the adjective precedes or follows the noun, b) whether the language uses pre- or postpositions, or c) whether in nominal genitive constructions the owner precedes what is owned or vice versa. Grammaticalization theory. Cycle of typological change syntactic > morphological > syntactic. LATs should comprehend language types. Structure of categories as they are stored in the brain; flexible in order for learning to take place. Side 10

NATO RESTRICTED Types of measurement Norwegian Defence Intelligence and Security School Intelligence Section What specific abilities would we like to measure for foreign language learning? o o o o o o o Ability to discriminate foreign language sounds phonetic decoding (Perception). Ability to mimic foreign (stretches of) sounds and intonation. Measurement of mimicry span and accuracy in reproducing unusual sounds. (Production) Does language ability correlate with musical ability? Not confirmed (1963) or only marginally so (Diller 1981) Ability to learn foreign language material by rote (Memory) Independent inductive learning: ability to develop meanings and rules inductively Flexibility to adapt to foreign language orthography (including to learn new alphabets) (memory, systematic processing, and production) Rate of learning Side 11

NATO RESTRICTED Norwegian Defence Intelligence and Security School Intelligence Section Language Aptitude Teaching methods and learning styles o o o learning style holistic or analytic learner. Two methods: Grammar-translation vs. active, practical use of the (spoken) language. Analytical method (traditional classroom teaching) vs. the audio-lingual method (e.g. in language labs). Matches or mismatches of the language teachers preferred teaching method. Ideally, LATs should capture both learning styles. Methods of selection aimed to predict success in language learning o o o Trial courses. Not efficient with respect to time and money. Artificial language test. Subjects were set to learn an artificial language vocabulary and certain grammatical principles, which were then to be translated into a natural language Phonetic script tests. Tend to correlate with judgments on phonetic mimicry ability Side 12

NATO RESTRICTED Norwegian Defence Intelligence and Security School Intelligence Section Challenges in presently taught languages o o We need to measure analytical skills, i.e the ability to internalise and use grammatical rules Need to know whether students are apt to learn language structures very different from their own mother tongue o Presently relevant languages include Arabic, Russian, Pashto, and Dari Future potentially languages of interest could include Chinese and a number of African languages, French Haitian Kreyol Phonemic inventories with unfamiliar sounds and phonemic distinctions to which the language learned is unaccustomed. How do we make 2nd language learners aware of new and challenging phonetic distinctions? Do we need to unlearn our own language? Important distinctions to be aware of for Norwegian learners:

Voiced vs unvoiced sibilants (e.g. s z) Vowel length ([a] vs. [a:]) Velar and pharyngeal fricatives and stops Palatalized vs. non-palatalized consonants with same place of articulation NOTE: faulty pronunciation may distort meaning with fatal consequences. Side 13 NATO RESTRICTED Norwegian Defence Intelligence and Security School Intelligence Section What grammatical challenges do we encounter in these languages? o o Complex verbal inflectional paradigms Unusual morphological sequencing o Derivational morphology not present in Norwegian o Present in Persian and Pashto (common also in Romance languages)

SOV word order o Causative Directional morphemes (towards speaker, away from speaker) Pro-drop (in case of redundancy grammar) o infix as opposed to suffix or prefix (e.g. future particle in Pashto breaks up the verb, intensifying morpheme/causative in Arabic geminates the middle root consonant, broken plural of nouns in Arabic) circumfix or sandwiching prepositions: PP-NP-PP Additional complication in e.g. Pashto: in more complex noun phrases with genitives, adjectives and articles, where does the subject end and where does the object start? Unusual agreement phenomena Verb object agreement (Pashto verbs agree with direct object in past tense because of ergative sentence structure) Verb subject agreement Agreement within the noun phrase Agreement with relative pronoun Side 14 NATO RESTRICTED LATs used in the US

MLAB Modern Language Aptitude Test Norwegian Defence Intelligence and Security School Intelligence Section o o Research led by John Carroll and Stanley Sapon. Developed in the 1950s DLAB - Defense Language Aptitude Battery o Developed in the 1970s by Dr. Tony Al-Haik and Col Kebbey Horne. (For further details, see article Petersen, C.R, and Al-Haik, A.R. 1976. The Development of the Defense Language Aptitude Battery (DLAB). Educational and Pshychological measurement, 36. Pp. 369-380.) DLAB was based on research tradition on the MLAT work led by Carroll and Sapon. Presently used for selection and classification of language students at DLI Foreign Language Center. Replaced ALAT Army Language Aptitude Test and DLAT 1 and 2. o Structure resembles Eastern Europe languages. High predictive power for all languages in terms of success in language courses o Multiple choice. 2 hour test. Consists of four subparts:

1) Biographical inventory: Past experiences with language learning; attitudes for learning languages; academic merits 2) Phonology: recognition of stress patterns. Among four words with equal number of syllables, pick out the one with the aberrant stress pattern. 3) Foreign language grammar. Artificial language material is presented along with grammatical rules, both syntactic and morphological. Examinees task is to apply these rules to new language material. Tasks have cumulative difficulty. 4) Concept formation. Four pictures, each with artificial language texts. This gives the foundation for the forming of concepts. Three additional pictures are presented, accompanied by four new artificial language texts. The task consists of correctly matching texts with pictures. Side 15 NATO RESTRICTED A brief State of the Art of LATs? Norwegian Defence Intelligence and Security School Intelligence Section Success and characteristics of various LATs o o In the past the MLAT test has been used in the US with quite some success, e.g. in French. However, this test is very much based on language structures in Western Indo-European languages, i.a. SVO word order and not too much morphology. This is according to Parry and Stanfield (1990), I cannot confirm since I have not seen the test myself. The VORD test seeks to remedy this bias. This is an artificial language (constructed language) test, made to resemble the structure of Turkic languages:

o Agglutinating: one morpheme (affix) one meaning. Clear morpheme boundaries. SOV word order Again, according to Parry and Stanfield (1990), the MLAT and US Army LA tests did not have any success with predicting language aptitude for languages with grammatical structures very different from Western IndoEuropean languages (e.g. Japanese, Korean, and Vietnamese). It seems therefore that tests like VORD are better suited for many of the languages we need in International Operations today, which means that being able to master unusual language structures is crucial. Side 16 NATO RESTRICTED Norwegian Defence Intelligence and Security School Intelligence Section In general, four abilities should be tested o 1) Memory abilities the ability to learn a large number of semantic-symbol and/or sound-symbol associations in a relatively short time period. o Linguistic rote learning ability (not general memory) 2) Phonetic coding ability the ability to identify and store new language sounds or stretches of sounds in long term memory

E.g. discrimination of tonemes i.e. syllabic tone patterns which distinguish between lexical or grammatical meanings (Chinese has 5, Hausa 3, Norwegian 2). Mimicking sounds accurately. Being able to learn spelling and segmentation of foreign words. Recognition of phonemic features which are not distinctive in ones own mother tongue Side 17 Norwegian Defence Intelligence and Security School Intelligence Section NATO RESTRICTED o 3) Grammatical sensitivity the individuals ability to demonstrate an awareness of syntactic patterning of sentences in a language and the grammatical functions of the individual elements in a sentence. o Correlates with knowledge of grammatical concepts and terminology May be tested on the basis of analogy; examinee is asked to find a grammatical construction that has a function analogous to that of an indicated word/phrase in a sentence. 4) Inductive learning ability the ability to infer forms, rules and patterns from systematically varying language material (= factor analysis)

Here language aptitude is most closely related to general intelligence Method: present materials in an artificial language in such a way that learners will use this ability in learning the foreign language. The test as a learning task Side 18 NATO RESTRICTED Predictive power of LATs Norwegian Defence Intelligence and Security School Intelligence Section One might not be able to include all types of linguistic rules in a LAT Do we have predictive power with only a subset of types of grammatical rules? o o It is hoped that while measuring these 4 skills, one will be able to predict language aptitude also for languages which are characterized by a slightly different set of rules. I.e. what one measures is not the specific rule type, but the apprehension and application of rules per se. Cf. the DLAB This may be confirmed or not on the basis of statistical data on pass rates in pragmatic terms (does the LAT do the job?), but notice that many variables are difficult to isolate if this is to be a scientific conclusion (e.g. how does one measure the influence of

motivation, prior knowledge, learning styles and teaching methods?) Side 19 NATO RESTRICTED Language equals concept formation Norwegian Defence Intelligence and Security School Intelligence Section Information packaging in the clause o o o Languages package grammatical information differently. Part of the job of learning a new language is to become aware of the new types of grammatical content that may now be expressed in that language, and that certain distinctions that you did not have to consider in your own language, now is obligatory. Ferdinand de Saussure talks about meaning as an amorphous mass. Prior to the influence of language, this amorphous mass is not sectioned into chunks of meaning. However, in the process of conventionalizing a sound-meaning relationship in the linguistic symbol, the individual languages impose structure on meaning. Languages do this differently, creating slightly different conventionalized grammatical concepts. This is why the ability of concept formation in a very true sense should be heeded when testing language aptitude, as well as when learning a new language. Side 20 NATO RESTRICTED

Norwegian Defence Intelligence and Security School Intelligence Section Summary We have looked at: o o o o o o o o The concept of Language Aptitude and what it is not (motivation, exposure to language, etc.) Aptitude is not achievement but can be confused in practical terms What do we need to measure? What is the language faculty like and sources of evidence for assigning it to a special faculty or to general cognitive abilities (not conclusive) What LATs are presently in use and what is their outline? What components should a good LAT include? How do we know they have predictive power? Why concept formation ability is highly relevant to language aptitude Side 21

Recently Viewed Presentations

  • Comparative and Superlative Adjectives

    Comparative and Superlative Adjectives

    If an adjective ends in E, this is removed when adding ER or EST. If an adjective ends in a consonant followed by y, y is replaced by i when adding ER and EST. PowerPoint 프레젠테이션 PowerPoint 프레젠테이션 PowerPoint 프레젠테이션...
  • Training


    Operator Overloading What is operator overloading? Most predefined operators (arithmetic, logic, etc.) in C++ can be overloaded when applied to objects of a class by providing an operator


    RESEARCHING MULTICULTURAL BRITAIN PROF JOHN EADE CRONEM ROEHAMPTON/SURREY Reformism and Empirical Research: 1850-1945 Victorian pioneers - Mayhew, Booth and empirical study of urban social and economic inequality The Labour Party - Fabians, municipal socialism, LSE, Mass Observation, sociology and anthropology...
  • Building Effective Leadership Teams How can independent schools

    Building Effective Leadership Teams How can independent schools

    rom Research and Practice (American Council on Education, 2017) ... unable to show vulnerability or weakness. Teams overcome this dysfunction by sharing experiences, developing insight into the unique characteristics of other team members. 2. Fear of
  • Big Data Mining (巨量資料探勘) - Tamkang University

    Big Data Mining (巨量資料探勘) - Tamkang University

    1 2017/02/16 巨量資料探勘課程介紹 (Course Orientation for Big Data Mining) 2 2017/02/23 巨量資料基礎:MapReduce典範、Hadoop與Spark生態系統 (Fundamental Big Data: MapReduce Paradigm, Hadoop and Spark Ecosystem)
  • The Best Bones Possible - Michael E. Greer, MD

    The Best Bones Possible - Michael E. Greer, MD

    Ostroff LE, Cain CK, Jindal N, Dar N, Ledoux JE. Stability of presynaptic vesicle pools and changes in synapse morphology in the amygdala following fear learning in adult rats. J Comp Neurol. 2011 Jun 14
  • Welcome to ITRC's Internet Training - CLU-IN

    Welcome to ITRC's Internet Training - CLU-IN

    The optimization should be focused on finding ways to reduce the effort in those areas where the operators spend most of their time. The slide lists some alternatives for reducing labor. Any change in the system that would have a...
  • Directors Report Unidata Policy Committee Meeting 24 May

    Directors Report Unidata Policy Committee Meeting 24 May

    Director's Report Unidata Policy Committee Meeting 24 May 2004 Mohan Ramamurthy Unidata Program Center UCAR Office of Programs Boulder, CO