Spring 2001 • ENGLISH 110J • Professor Tanaka

CT MODEL 3

BASIC ANALYSIS AND EVALUATION

 AT ANY RATE, THESE CRITICAL THINKING MODELS HAVE COME TO DEFINE WHAT WE MEAN BY A "LOGICAL" OR "REASONABLE EXPLANATION" IN EVERYDAY LIFE. HOWEVER, JUST AS IN THE CASE OF "DEFINING" STANDARD ENGLISH, THERE IS NO SINGLE CRITICAL THINKING TEXT OR AUTHORITY THAT EVERYONE HAS AGREED IS "CORRECT" OR "TRUE FOR EVERYONE." NONETHELESS, THE SUBJECT CONTINUES TO BE TAUGHT AND THE CONCEPT CONTINUES TO BE USEFUL.

OF COURSE, "OBJECTIVE," LOGIC-BASED EXPLANATIONS ARE NOT THE ONLY KINDS WE GIVE IN COMMUNICATING OUR ORDINARY THOUGHTS AND BELIEFS TO OTHERS. FOR EXAMPLE, WE MIGHT PRESENT MORAL, RELIGIOUS OR EMOTIONAL EXPLANATIONS AS WELL. HOWEVER, THESE DO NOT HAVE TO FOLLOW THE SAME KINDS OF CRITICAL THINKING RULES. BY THE SAME TOKEN, THEY ARE MORE DIFFICULT FOR THOSE WHO DO NOT THINK AS WE DO TO SHARE AND EMPATHIZE WITH.

CRITICAL THINKING BECOMES VITALLY IMPORTANT WHEN WE WANT TO ESTABLISH A SHARED BELIEF. IF WE BELIEVE THAT X IS TRUE OF THE WORLD <THAT IT IS A "FACT"> AND WE WANT SOMEONE ELSE TO ALSO BELIEVE THAT X IS TRUE OF THE WORLD <OR AT LEAST ACKNOWLEDGE THAT WE HAVE GOOD REASONS FOR OUR TAKING IT AS A FACT>, THEN WE MIGHT TRY TO GIVE THEM AN EXPLANATION.

BUT TO EXPLANATION WHAT EXPLANATIONS ARE, WE HAVE TO BACK UP A BIT AND TALK ABOUT THEORIES.

 THEORIES AND EXPLANATIONS

  LET US SAY THAT X IS SOMETHING THAT EXISTS IN THE WORLD.  THAT US SAY THAT AN EXPLANATION OF X TELLS WHAT X IS AND WHY IT IS FOLLOWING A MORE GENERAL SET OF RULES AND DEFINITIONS.  THE USUAL EXPLANATION CONSISTS OF ONE GENERAL STATEMENT AND ONE SPECIFIC FACT STATEMENT.

 THERE ARE MANY DIFFERENT KINDS OF EXPLANATIONS DEPENDING UPON SPECIFIC SITUATIONS AND NEEDS.

 LET US SAY THAT THERE ARE FORMAL EXPLANATIONS THAT ARE DERIVED FROM SETS OF RULES THAT HAVE NOTHING TO DO WITH THE REAL WORLD.

 THERE ARE ALSO SCIENTIFIC EXPLANATIONS THAT ARE MADE UP OF BOTH FORMAL RULES AS WELL AS RULES BASED UPON OBSERVATIONS OF THE PHYSICAL WORLD.  HENCE, SCIENTIFIC EXPLANATIONS ARE ALWAYS CONTINGENT.

 FINALLY, LET US SAY THAT THERE ARE INFORMAL EXPLANATIONS THAT ARE NEITHER STRICTLY LOGICAL NOR STRICTLY SCIENTIFIC BUT DO TRY TO FOLLOW THE RULES FOR CRITICAL THINKING AND DO TRY TO BASE THEIR ASSUMPTIONS ON OBJECTIVE FACTS AND OBSERVATIONS.

 MANY OF THE EXPLANATIONS WE ARE GIVEN IN TEXT BOOKS ARE INFORMAL IN THIS SENSE.  HOWEVER, MOST OF THEM ARE SIMPLIFIED VERSIONS OF FORMAL AND SCIENTIFIC EXPLANATIONS.

IN OUR DISCUSSION OF LANGUAGE IN 110J, WE WILL BE TALKING IN A STRICTLY INFORMAL SENSE ASSUMING THAT A SERIOUS, PROFESSIONAL  DESCRIPTION AND EXPLANATION OF GRAMMAR WOULD FALL IN THE DOMAIN OF LINGUISTICS.  IN MOST CASES, GRAMMATICAL PROBLEMS WOULD BE EXPLAINED BY SCIENTIFIC MODELS IN SYNTAX AND SEMANTICS.

HOWEVER, 110J IS NOT A LINGUISTICS COURSE, EVEN THOUGH IT IS A COLLEGE LEVEL COURSE AND HENCE MUST UTILIZE THE GENERAL CT MODEL UNDERLYING ALL COLLEGE LEVEL ANALYSIS.

THEREFORE, WHAT WE WILL BE DOING  IS TRYING TO COMPARE THE EXPLANATIONS GIVEN IN OUR TRADITIONAL GRAMMAR TEXTS WITH THE GUIDELINES FOR GENERATING INFORMAL EXPLANATIONS, OUR CT MODEL.

1. EXPLANATIONS

OK. LET’S GO BACK TO THE PARAGRAPH I USED TO OPEN THIS SECTION. LET US SAY THAT X IS SOMETHING THAT EXISTS IN THE WORLD.  THAT US SAY THAT AN EXPLANATION OF X TELLS WHAT X IS AND WHY IT IS FOLLOWING A MORE GENERAL SET OF RULES AND DEFINITIONS. 

SO IF X HAPPENS TO BE MODERN CLASSROOM ENGLISH AS DISCUSSED BY MODERN TRADITIONAL GRAMMAR (MTG) TEXTS, THEN AN EXPLANATION OF X WOULD INCLUDE ALL OF THOSE EXPLANATIONS OF ENGLISH PROVIDED BY, SAY, GRAMMAR SMART AND EHRLICH AND MURPHY’S ENGLISH GRAMMAR.

THE USUAL EXPLANATION CONSISTS OF ONE GENERAL STATEMENT AND ONE SPECIFIC FACT STATEMENT.

        FOR EXAMPLE:

        1. ONLY ADJECTIVES MODIFY NOUNS.

        2. ‘APPLE’ IS A NOUN.

        3. ‘RED’ IN ‘RED APPLE’ MODIFIES ‘APPLE.’

        4. THEREFORE, ‘RED’ IN ‘RED APPLE’ IS AN ADJECTIVE.

NOTE THAT THE TWO SPECIFIC FACT STATEMENTS, 2 AND 3, ARE ALSO DERIVED FROM MORE GENERAL DEFINITIONS: WHAT A NOUN IS AND WHAT A MODIFIER IS. THE GENERAL STATEMENT DEFINING ‘ADJECTIVE’ IS ALSO DEPENDENT UPON THE CONCEPT OF ‘NOUN’ AND ‘MODIFIER.’

WHERE DO THESE GENERAL RULES AND OPERATIONS THAT ARE USED TO MAKE EXPLANATIONS COME FROM?  THEY COME FROM SETS OF RULES, OPERATIONS AND OTHER ASSUMPTIONS CALLED THEORIES.

2. THEORIES

A THEORY IS THE SET OF GENERAL RULES (WHICH INCLUDE DEFINITIONS) AND OPERATIONS THAT ALLOW US TO MAKE EXPLANATIONS.  HENCE, ANY EXPLANATION OF X PRESUPPOSES A THEORY OF X.

IN A UNIVERSITY ENVIRONMENT, ANY THEORY OF X AND ANY EXPLANATION GENERATED BY THAT THEORY MUST, AS A MINIMUM, MEET THE BASIC RULES FOR CRITICAL THINKING.

FOLLOWING OUR SIMPLIFIED MODEL, LET US SAY THAT LIKE EXPLANATIONS, THEORIES CAN BE FORMAL, SCIENTIFIC OR INFORMAL.

SINCE WE HAVE SAID THAT MTG EXPLANATIONS ARE INFORMAL, LET US ALSO SAY THAT MTG EXPLANATIONS ARE ALSO INFORMAL.  IN FACT, THEY ARE SO INFORMAL THAT MOST WRITERS OF MTG TEXTS DO NOT EVEN SPELL OUT MANY OF THEIR GENERAL ASSUMPTIONS AND RULES AT ALL.  THEY ARE IMPLICITLY IMPLIED OR SIMPLY ASSUMED TO BE COMMON  KNOWLEDGE.

THE REASON FOR IS THAT THE BASIC FORMAT OF MTG MODELS IS PRESCRIPTIVE OR NORMATIVE.  THEY WERE DEVELOPED EVEN BEFORE THE PROCEDURES FOR PRESENTING LOGICALLY AND EMPIRICALLY VALID ARGUMENTS WERE EVEN FORMULATED OR UNIVERSALIZED AS THEY ARE NOW.  THIS IS NOT A TRICK OR ANYTHING.  110J IS CALLED TRADITIONAL GRAMMAR.  HENCE, ALMOST BY DEFINITION  THE MODEL OR MODELS  THAT REPRESENT THE SUBJECT ARE NOT LOGICALLY RIGOROUS.  THIS DOES NOT MEAN THAT THEY ARE NECESSARILY ILLOGICAL.  RATHER, LOGICAL CONSISTENCY AND OTHER RULES OF GOOD  ARGUMENTATION ARE NOT THEIR PRIMARY CONCERN.

SO THIS IS THE POSITION YOU ARE IN:

        110J STUDENT                          110A STUDENT

        MTG MODELS                              SCIENTIFIC LINGUISTIC

        THAT ARE NORMATIVE                 MODELS THAT ARE

        AND NON-LOGICAL.                     EMPIRICAL, EXPLANATORY

        SAMPLE TEXT:                             AND ASSUME A FORMAL,

        GRAMMAR SMART.                       LOGICALLY RIGOROUS                                                                         THEORY.

                                                        SAMPLE TEXT:

                                                        RODMAN AND FROMKIN,

                                                        LANGUAGE.

BUT EVEN THOUGH YOU, AS A STUDENT IN 110J, CANNOT REALLY CRITICIZE MTG’S FOR NOT BEING LINGUISTIC MODELS, YOU CAN CERTAINLY CRITIQUE THEM FROM THE POINT OF VIEW OF OUR GENERAL CT MODEL.

PRELIMINARY ASSUMPTIONS

FIRST OF ALL, I SOMETIMES USE THEORY AND MODEL INTERCHANGEABLY, ESPECIALLY WHEN TALKING ABOUT MTG THEORIES.  GENERALLY SPEAKING, WHEN A SET OF ASSUMPTIONS IS BEING TESTED, IT IS NOT ACCEPTED AS TRUTH.  IT IS A MODEL.  HOWEVER, A MODEL THAT IS THEN ACCEPTED AS TRUE, IS THEN GIVEN THE STATUS OF A THEORY.  IT’S RULES AND OPERATIONS ARE ASSUMED TO BE CORRECT.

ON THE OTHER HAND, THE RULES OF A GRAMMATICAL MODEL AND THE RULES OF A GRAMMATICAL THEORY ARE OFTEN TRYING TO DO THE SAME THING. HENCE, IN THE DISCUSSION TO FOLLOW, I’M JUST GOING TO SET UP SOME GENERAL RULES FOR THEORIES AND EVALUATING THEORIES.  BUT IT IS CLEAR THAT NO TRADITIONAL GRAMMAR OF ENGLISH CAN BE A THEORY <IN THE SCIENTIFIC OR FORMAL SENSE) OF ENGLISH.

NOW SINCE I AM BEING VERY GENERAL HERE, I WILL JUST SET OUT SOME BASIC RULES THAT ANY CANDIDATE FOR A THEORY OF X SHOULD MEET.  THEN I WILL GIVE SOME EVALUATION PROCEDURES FOR CHOOSING BETWEEN TWO POSSIBLE THEORIES. 

A THEORY OF X USUALLY CONSISTS OF A SET OF ELEMENTS AND RULES THAT DETERMINE THE RELATIONSHIP BETWEEN THESE ELEMENTS.

THIS MEANS A THEORY OF X SHOULD PROVIDE US WITH DEFINITIONS THAT IDENTIFY THE ELEMENTS AND CATEGORIES OF ELEMENTS THAT THE THEORY COVERS AND THEN A SET OF RULES FOR HOW THESE ELEMENTS AND CATEGORIES CAN RELATE TO EACH OTHER.

SO WE USUALLY START BY LOOKING FOR A SET OF DEFINITIONS.  THESE DEFINITIONS WILL DEFINE THE ELEMENTS COVERED BY A SET OF RULES AND OPERATIONS.  IN FORMAL GRAMMARS, WE USUALLY DISTINGUISH BETWEEN THREE DIFFERENT CATEGORIES OF LANGUAGE.

1) SYNTAX: DEALS WITH HOW WORDS ARE COMBINED TO FORM SENTENCES.

2) SEMANTICS: DEALS WITH THE MEANINGS OF WORDS AND SENTENCES.

3) PHONOLOGY: DEALS WITH THE SOUNDS OF WORDS AND

SENTENCES.

MTG’S USUALLY JUST BLEND SYNTAX, SEMANTICS AND COMMON USAGE, WHICH IS OFTEN A MATTER OF CHOICE AND STYLE.  SO SOME OF THESE RULES WILL BE DIFFICULT TO APPLY.

ALSO, THESE INFORMAL RULES ARE NOT COMPLETE.  HOWEVER, THEY ARE THE ONES THAT I BELIEVE ARE THE MOST USEFUL IN DOING THE KINDS OF INFORMAL ANALYSES AND EVALUATIONS WE WILL BE ATTEMPTING IN 110J.

•RULE 1

THE FIRST RULE FOR CONSTRUCTING A THEORY IS THE PRINCIPLE OF CONSISTENCY. THIS SIMPLY MEANS THAT ONE SHOULD NOT CONTRADICT ONESELF IN MAKING STATEMENTS ABOUT X OR OTHER ELEMENTS OF THE THEORY.

•RULE 2

THE SECOND RULE IS THAT THE DEFINITIONS IN A THEORY SHOULD HAVE SOME USE. THEY MUST DO WORK. IN OTHER WORDS, THEY SHOULD ENABLE US TO ACTUALLY IDENTIFY AND DISTINGUISH THE ELEMENTS OF X, THE SUBJECT THAT THE THEORY COVERS.

•RULE 3

THE THIRD RULE IS THAT THE CATEGORIES USED TO IDENTIFY AND GROUP THE ELEMENTS OF X SHOULD HAVE SOME DESCRIPTIVE CONTENT. THEY SHOULD ALSO HAVE A CLEAR LOGICAL RELATIONSHIP TO THE OTHER CATEGORIES AND GROUPS IN THE THEORY. 

FOR EXAMPLE, SCIENTISTS HAVE DETERMINED THAT A NATURAL LANGUAGE LIKE ENGLISH CAN BE EXAMINED BY LOOKING AT THREE GENERAL CATEGORIES OF FEATURES, SYNTAX, SEMANTICS AND PHONOLOGY.  THIS CATEGORIZATION IS BASED ON OBSERVATIONS OF HOW NATURAL LANGUAGES ACTUALLY WORK.  SECONDLY, THESE CATEGORIES ARE THEN BROKEN DOWN INTO MORE SPECIFIC SUBCATEGORIES. FOR EXAMPLE, PHONOLOGICAL RULES STATE THE UNDERLYING SOUND SYSTEM OF A LANGUAGE.  PHONETIC RULES DETERMINE HOW THOSE PHONOLOGICAL RULES WILL BE INTERPRETED TO PRODUCE SPEECH SOUNDS.

ALTHOUGH DEFINING CATEGORIES AND THE RELATIONSHIPS BETWEEN CATEGORIES IS AT SOME POINT ‘ARBITRARY,’ IN SCIENCE, ONE ALWAYS TRIES TO LOOK FOR "NATURAL CATEGORIES" OR GROUPINGS OF ELEMENTS BASED ON COMMON SETS OF "OBJECTIVE,” OBSERVABLE FEATURES. IN OTHER WORDS, IN A GOOD THEORY THE SYSTEM OF CATEGORIES AND GROUPS DEFINED BY THE THEORY SHOULD APPEAR TO FIT THE DATA IN A LOGICALLY CONSISTENT FASHION.

•RULE 4

A THEORY OF X IS SUPPOSED TO DO THREE THINGS: TO DESCRIBE X, TO EXPLAIN X AND TO MAKE PREDICTIONS COVERING X.  WHEN CONSTRUCTING A THEORY OF X, YOU WANT TO DEVELOP DEFINITIONS AND RULES THAT COVER AS MUCH OF X AS POSSIBLE.  IN OTHER WORDS, YOUR THEORY SHOULD ALLOW YOU TO GIVE CONSISTENT DESCRIPTIONS OF X AND MAKE CONSISTENT EXPLANATIONS AND PREDICTIONS THAT COVER X.

YOUR THEORY SHOULD DO THE WORK OF A THEORY.

EVALUATION CRITERIA

NO THEORY OF THE NATURAL WORLD IS PERFECT.  EVERY THEORY IS INADEQUATE IN ONE WAY OR ANOTHER.  SO EVERY THEORY IS, IN A SENSE, A MODEL.  BUT WHEN A MODEL  IS SIGNIFICANTLY BETTER THAN COMPETING MODELS, IT IS THEN USEFUL TO SIMPLY REGARD IT AS TRUE, REGARD IT AS A THEORY.

WHAT ARE SOME OF THE CRITERIA THAT WE CAN USE TO SAY THAT ONE THEORY OF X  MIGHT BE BETTER THAN OTHER THEORIES OF X? THE FOLLOWING ARE SOME HELPFUL GUIDELINES.

1. ANSWER OBJECTIONS

A GOOD THEORY OF X SHOULD NOT ONLY  OFFERS REASONS THAT EXPLAIN X, BUT IT SHOULD ALSO TAKE INTO ACCOUNT OF REASONABLE OBJECTIONS TO THE THEORY ITSELF. ARGUMENTS AND EXPLANATIONS ARE NEVER CREATED OR PRESENTED IN A VACUUM. HENCE, IN DEVELOPING A WORKABLE THEORY, ONE MUST ALWAYS TAKE INTO ACCOUNT OTHER POINTS OF VIEW AND OTHER WAYS OF INTERPRETING THE DATA.

OTHER THINGS BEING EQUAL, GIVEN TWO THEORIES, THE ONE THAT DOES THE BETTER JOB OF DEFENDING ITSELF IS STRONGER.

2. COMPLETENESS

WE SAID THE THREE PRIMARY FUNCTIONS OF A THEORY OF X ARE TO PROVIDE DESCRIPTIONS AND EXPLANATIONS OF X AND MAKE PREDICTIONS ABOUT X.  WE ALSO SAID THAT IN THE REAL WORLD, ALL COMPETING THEORIES OF X WILL BE INCOMPLETE.

THEREFORE, OTHER THINGS BEING EQUAL, IF A THEORY  IS ABLE TO GENERATE DESCRIPTIONS, EXPLANATIONS AND/OR PREDICTIONS COVERING MORE OF THE DATA THAN A COMPETING THEORY, THEN THAT THEORY IS THE STRONGER OF THE TWO.

IN OTHER WORDS, IF THEORY 1 CAN PROVIDE ADEQUATE DESCRIPTIONS BUT FAIL TO PROVIDE EXPLANATIONS AND PREDICTIONS OF THE DATA AND IF THEORY 2 CAN PROVIDE BOTH DESCRIPTIONS AND EXPLANATIONS  OF THE DATA (WHILE STILL FAILING TO GENERATE PREDICTIONS), THEN, OTHER THINGS BEING EQUAL, THEORY 2 IS STRONGER THAN THEORY 1.

3. ANSWERING QUESTIONS THAT MATTER

OF COURSE, IN THE REALM WORLD, NOT ALL DATA IS CREATED EQUAL.  LET’S SUPPOSE THAT TWO THEORIES ARE ABLE TO GENERATE ADEQUATE DESCRIPTIONS, EXPLANATIONS AND PREDICTIONS OF, SAY, 70% OF THE DATA OF X.  HOWEVER, THE RULES OF THEORY 1 DO NOT ALLOW IT TO EXPLAIN MANY OF THE MOST IMPORTANT OR SIGNIFICANT FEATURES OF THE DATA.  ON THE OTHER HAND, THE RULES OF THEORY 2, WHILE THEY ARE STILL INCOMPLETE, DO COVER THOSE FEATURES OF THE DATA THAT ARE IMPORTANT.

IN THIS CASE, WE WOULD SAY THAT THEORY 1 MAKES MORE "SIGNIFICANT GENERALIZATIONS" THAN THEORY 2.  IN OTHER WORDS, OUT OF ALL THE POSSIBLE STATEMENTS THAT CAN BE MADE ABOUT X, THEORY 2 ALLOWS ONE TO MAKE THE EXPLANATIONS AND PREDICTIONS THAT REALLY MATTER TO THOSE USING THE THEORY.

THEREFORE, OTHER THINGS BEING EQUAL, IF A THEORY  IS ABLE TO GENERATE MORE SIGNIFICANT GENERALIZATIONS COVERING THE DATA THAN A COMPETING THEORY, THEN THAT THEORY IS THE STRONGER OF THE TWO.

4. SIMPLICITY IS BETTER

FINALLY, THERE IS THE SIMPLICITY CRITERION. THIS STATES THAT IF TWO THEORIES MEET THAT ALL THE FORMAL CRITERIA FOR THEORIES, THEN , OTHER THINGS BEING EQUAL, THE ONE THAT IS THE SIMPLEST, THE ONE THAT INVOLVES THE FEWEST RULES AND ASSUMPTIONS AND REQUIRES THE FEWEST STEPS TO GENERATE EXPLANATIONS, IS THE STRONGER.

IN BOTH SCIENCE AND MATHEMATICS, ONE STRIVES TO GIVE PROOFS AND EXPLANATIONS THAT ARE "ELEGANT," THAT HAVE THE MOST EXPLANATORY POWER WITH THE LEAST AMOUNT OF WASTED "EFFORT."

SUMMARY

LET’S THROW ALL OF THESE RULES INTO A BLENDER AND COME OUT WITH THE FOLLOWING THREE GUIDELINES FOR WHAT WE ARE LOOKING FOR IN A GOOD MTG.  REMEMBERING THAT WE ARE NOT EXPECTING A TRADITIONAL GRAMMAR TO BE A THEORY IN  THE WAY I’VE OUTLINE ABOVE, I THINK WE CAN AT LEAST ASK IT TO DO THE FOLLOWING:

a) OFFER A CLEAR AND CONSISTENT SET OF DEFINITIONS AND RULES COVERING ENGLISH GRAMMAR. IT SHOULD TALK CLEARLY AND CONSISTENTLY ABOUT THE FEATURES OF ENGLISH THAT ARE DEMONSTRABLY SIGNIFICANT OR IMPORTANT.

b) GIVE WITH GOOD, CLEAN EXPLANATIONS OF REALLY TOUGH PROBLEMS.  THE REASON MOST PEOPLE LOOK TO A THEORY OR A GRAMMAR IS BECAUSE THEY HAVE A QUESTION.  A GOOD GRAMMAR SHOULD NOT ONLY COVER THE BASIC ELEMENTS, IT SHOULD GIVE CONSISTENT EXPLANATIONS OF PROBLEMATIC ISSUES.

c) ADDRESS OR TAKE INTO ACCOUNT VALID OBJECTIONS TO ITS OWN RULES AND ASSUMPTIONS OR AT LEAST A STATE ITS LOGICAL LIMITATIONS.

NEEDLESS TO SAY, THERE ARE FEW, IF ANY, MTG’S, THAT WILL MEET THE FIRST OF THESE CRITERIA, LET ALONE THE OTHER TWO.  BUT THIS IS WHAT WE WILL BE BRINGING TO THE DISSECTION TABLE THIS SEMESTER.

ONE OF THE TYPES OF MODELS YOU MIGHT LOOK OUT FOR ARE HYBRIDS.  THESE ARE MTG’S WRITTEN BY THOSE WHO ARE LINGUISTS OR HAVE A LINGUISTICS BACKGROUND.  THESE HYBRIDS TRY TO PRESENT AN ENGLISH GRAMMAR FOR THE GENERAL STUDENT OR THE BARNES AND NOBLE READER THAT COMBINES ASSUMPTIONS FROM BOTH TRADITIONAL AND SCIENTIFIC GRAMMARS.  IF YOU CAN FIND A HYBRID THAT DOESN’T RUN INTO A LOT OF PROBLEMS, PLEASE LET ME KNOW ASAP.