University of South Florida Fall 95
Diachronic Linguistics in the Classroom:
Sound Shifts and Cognate Recognition.
Seminar in Second Language Acquisition
University of South Florida Fall 95
Diachronic Linguistics in the Classroom: Sound Shifts and Cognate Recognition.
Introductory level students of German who are native or near-native English speakers may have heard that English is a Germanic language but they seldom know how that fact can be of value to them in their efforts to learn German. A cursory view of the two languages in their current state will lead one to think the two are quite foreign to one another and that they bear little resemblance in any aspects. A closer view from an historical perspective, however, will show that their close kinship has left traces of their shared lineage in a consistent and predictable pattern. This pattern has been detected and modeled by germanists for the interest of historians and other germanists but its use as a tool for language learning has, so far, been largely overlooked. Students of German are seldom given instruction of the historical development of German (which includes the history of the English language) until they have already achieved a measure of language mastery and communicative proficiency.
This study proposes to test the theory that demonstrating some of the consistent structural relationships between English and German will help the beginning level learner of modern German to better recognize cognates and near cognates - thereby increasing ‘guess-power’ in vocabulary acquisition. A rudimentary knowledge of the primary sound shifts of the German language, an historical consideration, can be used as a very basic code-breaking tool for deciphering New High German vocabulary by English speaking students. To the beginning student, this can be of value as a potent vocabulary learning strategy for future use as well as giving them a considerable headstart in German language learning in general. It also gives the instructor a much larger vocabulary base upon which to demonstrate grammatical principles. In excess of 200 German words can be seen to be cognate pairs of English words when their phonemes are shifted according to the relationships outlined in the code breaking tool. The code breaking tool simply delineates some of the more stable phoneme relationships currently found between English and German.
The germanists Wolff (1988) and Crawford (1993) advocate those aspects of diachronic linguistics which could best be incorporated into the FL classroom and they stress that knowledge of historical background will add interest to the class when students’ questions turn to matters best explained by the language’s evolutionary stages.
The question of whether diachronic linguistics is of interest to students of modern High German is not at issue. Since historical grounding explains mysterious rule exceptions and often clears up the seemingly inexplicable, it is assumed that it would be of interest to the student. Historical linguistics helps account for synchronic variability by illuminating the source of current discrepancies. While conceding that it is of interest, it remains to be seen whether such a background knowledge can be useful as a tool for future language acquisition.
Much of the prior work on diachronic linguistics in the early level German classroom (Wolff, 1988, Crawford, 1993, Clausing, 1987, Smith 1968) is concerned with those aspects of the history of Old High German, Middle High German, Old English, Middle English and New High German which are easily incorporated into the beginning level curricula and which serve to answer frequent questions, specifically about gender, verb inflection, word formation, umlaut/ablaut mutations, and other aspects of the complex relationship between German and English. While the aforementioned aspects would be of interest, their usefulness as a learning tool is doubtful. It is not clear how knowing the history of gender declension, for instance, helps a student to learn vocabulary, i.e. memorize articles along with the substantive (noun). It would require the student to have a prior knowledge of OHG or OE to be able to see the lineage and to make use of it in vocabulary acquisition. To know about optional -e case endings in the dative singular helps only to answer the question of why one book says "zu Haus" and another says "zu Hause" but it doesn’t "arm" (Wolff 1993) a student with a useful cognitive tool for the future.
There are selected aspects of the German/English (sibling) relationship that would not only be interesting but would indeed be useful tools with which students could "arm" themselves for future ease of vocabulary acquisition, most notably, the consonantal and vocalic sound shifts.
While this study seeks to find empirical data to support the idea that aspects of this sibling relation may function well in the onset of German language instruction, there have already been two germanists who have implemented a portion of this idea into their elementary German textbook "Neue Horizonte". In the first publication (1984), D. B. Dollenmayer and T.S. Hansen gave a very brief notation about the common origins of English and German in a kind of interlude between chapters 4 and 5. In addition to pointing out a few clear cognates, they also mention that some consonants have a regular pattern of change or "alternation" giving a very few examples such as German /ss/ and English /t/ as in Wasser =water and gross=great. In the most recent (1996) edition of their text, the authors have taken this brief statement one large step further by labeling the alternations as outcomes of consonant shift and more importantly, they suggest that students themselves attempt to guess the meaning of near cognates rather than just pointing out the phonemic relationships resulting from historical sound shifts.
IV. Historical Language Evolution
English and German today are seen as branches of the same language family. They are like cousins born to mutual grandparents. Proto-Germanic is the name scholars have given to the Germanic branch of a yet older language referred to as Indo-European (3,000 - 4,000 BC). At the time of Proto-Germanic (500 BC - AD 60), English and German were not yet separate languages. At the time of Old High German (700 - 1050), the two were still of one language but assorted dialects could be distinguished, two of which were very similar to one another and today, their names are combined by scholars and referred to as Anglo-Saxon. The Anglo-Saxon dialects of Old High German are also referred to as Old English. Although today’s English takes much of its lexicon from romance languages, a 1986 computer analysis showed that the one hundred most commonly used English words are all Anglo-Saxon in origin (McCrum, Cran & MacNeil 1986). Through the period of Middle English (AD 1150 - 1500) the branch that was to become modern English underwent the orthographic loss of the inflectional endings which it once shared with German, that is, those endings had already begun to disappear in spoken OE and it is the reflection of that loss which, when it began to show in the written word, became known as Middle English (McCrum, Cran & MacNeil 1986). While Old English was undergoing the loss of inflectional endings, it retained much of the phonetic features that were shared in Proto-Germanic. It is the early German language that underwent the phonetic rebellion from its elder in that the inflectional endings were maintained while the sound system went through major changes. As Stephen Clausing (1987)points out, in two millennia since the time of Old High German, English has barely budged from its roots in terms of its consonant system.
Throughout its evolution, English has certainly gone through major changes in other ways, especially in the loss of nearly all of the original inflectional patterns, but in terms of the sound system of the alphabet, German is the runaway cousin. It is a runaway that has left a trail, however. Linguistic historians are able to track the evolution of German through two major consonant shifts and one major vowel shift.
From Indo-European to Proto-Germanic all Germanic languages alike experienced the first consonantal sound shift, also known as Grimm’s law, observed and first noted by the German philologist, better known for his fairy tales, Jakob Grimm, but the second consonant sound shift, called Vernor’s Law was unique to High German leaving the Low German dialects to have in common a contrasting phonology. The later vowel shift was also unique to High German leaving some low Germanic dialects to share striking similarities in spite of current geographic separations. For example, comparing vocalic similarities between Dutch and Swiss German, it is evident that neither language joined High German in the vowel shift.
While the details of the trail left by New High German’s evolution is apt more for the Germanist with a prior language mastery, the result of this trail is a consistent, tangible relationship between today’s English and German that is indeed fit curriculum for the introductory level student.
V. Synchronic Linguistics and Cognate Pairing
It is less the history of the two languages which is a potent tool, than the resultant current relationship and its application in cognate pairing. The modern spelling system of the English language stabilized more than 300 years ago (Pyles, 1986) and it retains many features that are remnants of the heritage it shares with modern German. One aspect of the phonological relationship to be considered is whether a cognate pair can be recognized for its acoustic match or for its orthographic resemblance. The German word "oft" has a greater chance of being paired with its English counterpart "often" if the words are seen in print than if they are heard. As auditory input, the word "often" has a better chance of being falsely matched with the German "Ofen" meaning "oven". In addition to contextual clues, a brief rundown of the current orthographic correspondence between the two phonological systems will afford a language learner a greater source of comprehension techniques.
The fact that there are a great many full cognates of German and English with no orthographic distinctions between them is immediately evident even to the introductory level student of German. The inventory of identical words shared by English and German has been estimated at 8000 (P. Braun 1985, 1979).
It is necessary to realize that a considerable amount of
English lexical items are not of Germanic origin as shown by Maria Martinez’
(1994) study of cognate recognition strategies in reading comprehension in which
two thirds of the 257 words in a subtechnical vocabulary sample were found to
be English/Spanish cognates. However, the words that English does have in common
with German are typically the oldest words of the sort that represent very universal
and basic concepts: father (Vater), mother (Mutter), brother (Bruder), home
(Heim), fire (Feuer), earth (Erde), water (Wasser), etc. Such words are likely
to have a high frequency of use. So even though these sibling words are fewer
in number, their recurrence in everyday speech gives them a greater importance
and makes recognizing them easily a desirable effect of this limited code breaking
VI. Research Question
This study begins with the proposal that a small amount of information concerning this historical phenomenon will indeed "arm" students and prepare them to recognize an additional host of partial cognates as very close relatives of English words. In fact, thorough knowledge of the first set of shifts (or a simple one page chart in hand) will render a great many German words practically ‘the same’. For example, knowledge that /t/ corresponds to /d/ initially and that /g/ relates to /y/ in final position renders the word "Tag" into ‘day’ in a systematic logical way, quite easily ‘guessed’ by a cognitive process.
Through the use of cognate instruction this study will not only delve into the English-German cognate relationship but through the phoneme relationship key, an attempt will be made to expand the possible connections that make cognate recognition possible.
V. Relevant Literature Review
The main impetus for this study comes from the work of Arteaga and Herschenson (1995). In their experimental study of diachronic linguistics in the first year French classroom, they tested whether instruction of those historical phenomena which have resulted in stable orthographic relationships can serve as an orthographic code breaker and enable an understanding of lexical patterning. They found that such knowledge indeed helps to increase guess-power in the acquisition of new vocabulary.
Several studies about reading in L1 have shown the importance of vocabulary knowledge. Hancin-Bhatt & Nagy (1993) point out that research on reading ability in L1 (Anderson & Freebody, 1981) shows that poor vocabulary knowledge has a negative effect on reading and subsequent poor reading has been shown to have a negative impact upon academic success (Garcia, 1988, 1991; Saville-Troike, 1984). In his commentary on free voluntary reading, Stephen Krashen (1995) joins others in emphasizing the effect of reading on other areas of L2 proficiency. Studies undertaken in French (Treville 1993) and Spanish (Nagy 1992, Hancin-Bhatt & Nagy 1993) show the importance of cognate recognition in reading comprehension and more importantly, the awareness of cognate relationships in reading strategies (Nagy, Garcia, Durgunoglu & Hancin-Bhatt 1991). The researchers in the aforementioned study concluded that explicit instruction of cognate pairing will increase learners’ utilization of cognate knowledge. They consider such instruction necessary because their studies consistently found underuse of cognate pairing strategies for all age groups. These findings underscored previous data uncovered by Garcia (in 1988) that showed Spanish and English bilingual 5th and 6th grade students were failing to make use of obvious cognate relationships. Later, in 1991, Garcia (w/Pearson and Jimenez) reported that 6th and 7th grade Latino bilinguals who were proficient readers did make frequent and effective use of their cognate knowledge.
In the romance languages, French and Spanish, detailed studies have been conducted to access developmental patterns of cognate recognition abilities distributed along an age-group continuum. (Hancin-Bhatt & Nagy 1993, Treville 1993) In interviews with bilingual 5th and 6th graders, Hancin-Bhatt & Nagy (1993) compared the cognate recognition abilities of different grade levels in search of a developmental trend. They expected to find a trend based on maturity levels from pre-adolescents through adult cognition based on Inhelder, Piaget (1958) and Vygotsky’s (1962) claims that adult maturity allows a greater level of metalinguistic awareness and a sensitivity to linguistic structures. For students between ages 9 and 13 they did indeed find a rapid increase in Spanish L2 students’ ability to recognize cognates. Older learners exploited systematic relationships. Pre-teen "upper elementary bilinguals" relied on simple orthographic overlap and not on systematic suffix patterns. So although Hancin-Bhatt & Nagy claim that word recognition and vocabulary building are enhanced by the ability to analyze the morphological structure of words, their research shows that it takes a few years more experience with the language at increasingly higher levels of discourse to begin to have a ‘feel’ for the morphological relationships among words and wordfields. It doesn’t seem surprising that Hancin-Bhatt & Nagy found a jump in learners cognate recognition abilities, which they read as an increase in their metalinguisitic awareness as well, at just the stage in teen-aged life when student’s are forming both their individual identities (practicing their signatures for instance) and their group identities, which typically involves a new crop of slang entries into the language. These new slang words are related to existing words in the language. Carroll, (1992) would say that they are formally (morphologically or phonetically) related to existing entries in the lexicon and a seemingly new word or new grammatical categorization may be understood because it activates (automatically) an existing field of lexical entries.
There have also been studies that suggest that a transfer L1>L2 of suffix rules, when it occurs can facilitate recognition of words that do not have cognate stems. Adjemian’s (1983) experimental study produced data to show that learner’s hypotheses about the L2’s lexical rules and word formation processes are constrained by L1 rules and processes. From that it followed that Hahn’s study of Korean and Spanish adult learners of English would show that the more similarity in the structure of morphological rules between language pairs, the broader the possibility for cognate recognition in the L2. The Spanish learners were able to pick out more cognate pairs than were the Korean learners.
Hahn’s results suggest that recognition of affix systems can be higher in less similar language pairs. Because of the strong dissimilarity between languages, the Korean learners couldn’t recognize as many cognate stem morphemes, but they were able to make better use of affix patterns. Perhaps as a result of not having stem cognates at their disposal the alternate strategy was more heavily relied upon. This suggests that the Spanish learners did not make full use of cognate recognition strategies. Hancin-Bhatt & Nagy complain that while the older teen-aged learners had substantially more cognate recognition abilities than the pre-teenagers, they still fell far short of utilizing all potential recognition strategies. In other words, their scores were much better, but still dreadful. Hahn’s study shows that even with increased awareness of linguistic structure and morpheme relationships, adults still fail to take full advantage of cognate pairing. Courses of language study are sometimes introduced with a brief mention of cognate pairs or the first few nouns taught in the target language may be cognates phonetically but not orthographically: "Hier ist ein Buch". Students are seldom given detailed assistance in expanded cognate recognition beyond being told what the word "cognate" means. Perhaps, in time, it will be pointed out that some suffixes have a consistent correspondence between L1 and 2 "und das ist fantastisch," but it is seldom mentioned until experience has already shown it to be apparent.
As Arteaga and Herschenson conclude, the process that leads to cognate recognition is not automatic. It is novel to learners with no prior second language learning experience. Along with their (1995) work, the present study will show that such a process can be taught simply and briefly, yet yield results of substantial gain for the student.
VII. Aspects of Cognate Pairing
There is much to consider in determining which aspects constitute a cognate pair. In her article "On Cognates" (1992), Susanne Carroll claims that the Cohort Model (Marslen-Wilson, 1989) (Marslen-Wilson & Tyler, 1980) (Marslen-Wilson & Welsch, 1978) explains lexical activation and cognate recognition. The model is a structural-representational view of cognition upon which Carroll bases her view of cognate pairing not as "a special sort of behavior exhibited only by bilinguals, but rather a particular instance of a more general phenomenon." It posits the lexicon as the central link in language processing (Marslen-Wilson 1989) and as the central link in language learning as well. Carroll claims therefore, that it should be just as primary to second language acquisition. The mental lexicon is organized into informational units called "lexical entries". Each entry is associated with a "recognition element" also called an "address". This address is a "representation which can be activated by a detection mechanism examining input stimuli." Addresses may be arranged in "neighborhoods" which are based on formal similarities (e.g. phonological features or stem word semantics) and addresses are distributed among and between neighborhoods. In other words, they overlap much like word fields. During language comprehension, the input will automatically activate multiple addresses or recognition elements in the L1. Contextual clues and the processing of semantic and morpho-syntactic properties immediately begin reducing the number of addresses until a single entry is selected (or until a null value is chosen). According to Marslen-Wilson, the process occurs within 200 milliseconds from the initial word input. The end result is based on formal similarity - acoustic-phonetic features for auditory input and orthographic similarities for visual (written) input, and according to Carroll, it need not be correct in meaning to be considered a cognate match. Criteria for cognate pairing is based on whatever address is activated in the mental lexicon regardless of the etymological relationship (or lack thereof) between entry and input.
Transfer for orthographic processing has been shown to be a better facilitator of cognate recognition. In 1988, Rolf Palberg conducted five experiments with Swedish speaking students of English. His first and third experiments dealt with uninstructed students who were just beginning English courses at the ages of 9-10 years old. In the first experiment of the study the students were tested for recognition of English words based on aural input. It was concluded that cognate recognition played a major part in the students’ successful strategies because the scores were deemed highest for words which were "similar or almost similar in pronunciation to their Swedish translational equivalents", that is, they most often recognized the cognates that sounded the same. Such words may share phonological features while remaining orthographically distant. In the third experiment of Palberg’s study, uninstructed students were tested on their visual cognate recognition ability, that is, in a ‘think-aloud activity’ they were asked to read an English text for meaning by which means their decoding abilities were accessed. The text had been prepared with words "orthographically identical or nearly identical to their Swedish equivalents". The researcher attributes the greater part of their success at decoding the text meaning to the orthographic ‘sameness’ of cognate pairs.
It has been claimed that cognate pairing techniques can have a detrimental effect as well. Carroll points out that cognate recognition can actually be a hindrance to learning because of the error that can arise through false friend cognates. The pairing process doesn’t care about end meaning. In the case of false friends the cognate pairing process can have lamentable results. However, it should be noted that there is not an endless abundance of them and those that an L2 learner may attempt to use falsely will generally be noticed instantly and pointed out as wrong. The difference is exceptional and it is most likely marked for better memory retention. Therefore, while false friends are initially a hindrance to communication, their markedness makes them easier to notice, performing a service beneficial to learning. If you posit that cognate pairing is helpful even in the case of false or accidental cognates, then any increase in matching ability is desirable.
The subjects of the present study are 21 volunteer pre-introductory German students enrolled in university introductory German courses who have no prior experience with the German language beyond knowing that "Gesundheit" and "Kindergarten" are words of German origin. Between a pre-test and the identical post-test, subjects received information about the phono-orthological relationship between English and German and instruction on how to use the vowel and consonant relationship key. (written by Caplan, 1995 based on a design by K. Schmidt, 1974) The instructions were scripted on a one page handout so that in the case of study replication, the instruction portion of the procedure could be duplicated without variance. Twenty minutes were allowed for each test.
The pre-test consisted of lists of German words that are historical origins of the modern English cognate counterparts. A total of 173 words was used to determine the effectiveness of the relationship key. The instructions consisted of a single page document delineating the current orthographic relationship between some German and English phonemes, called the ‘letter relationship key’. This was distributed along with a one page document containing a brief discussion of the shared history of the two languages and a few tips for using this ‘code breaking’ tool. The post-test was identical to the first text except that each word was followed by a number indicating the amount of changes or shifts necessary to derive the English word.
After completion of the pre-test, respondents were invited to mention any prior German language experience they might have had. They were asked to add this information on the back of the test form. This was merely to determine the subjects’ prior experience with second language learning. This was not to weed out or separate those subjects with prior knowledge, but it was used to try to account for any wide discrepancies among the pre-test scores. Based on the findings of Hancin-Bhatt & Nagy (1993), it is predictable that even those with a prior knowledge of the existence of cognates, would fail to make substantial use of that awareness as a vocabulary learning strategy.
The phoneme pair recognition code to be used in this study is similar to that devised by Karl Schmidt, (see also Wolff) based on the historical stages of this ‘sibling distancing’. One concern to be addressed is that cognate pairing is not the majority rule, that is, it is by far the lesser, not the greater part of the modern German vocabulary meanings that may be surmised in this way. Since there is, however, a significant amount, this indicates a possible value in sharing this information with beginning students but it must be stressed or better yet demonstrated that this near cognate ‘guessability’ has only a limited application. It may be a fine-lined delicate balance between the possible benefits of this diachronic information and the equally possible dangers of this ‘little bit of knowledge’. Therefore, a follow up study would be to determine the effects manifested in learners when a list of German words does not share the characteristics of the near-cognate sibling words. Such a study would probably look at those same subjects at the end or middle of the introductory German course to assess their vocabulary acquisition thus far (active and passive). The systematic relationship between English and German may be used to show how closely akin those languages are and that acquisition of the ‘foreign cousin’ language is within reach. At this point, however, it is merely the extent to which cognate recognition can be increased using such a code that was the goal of the current study.
XI. Analysis of Findings
A. Change in the Number of Answers Attempted. (table 1)
On the pretest, the number of attempts totaled for all 21 students was 1050. That number increased to 1240 on the posttest, which represents an overall increase of 18% in the total number of attempts. Statistically, this is signifigant at .10 probability.
Attempts for the first page alone went from 532 on the pretest up to 1045 total on the posttest. Since the test was three pages long and student’s attempts increased on the first page, It then follows that the number of attempts for the second and third pages would descend drastically; from 429 for page 2 pretest down to 188 posttest or down 56%. Attempts for page 3 showed an even steeper decline from 89 the number slides down to 7 for a total decrease of 92%
Table 3 shows this in the slight rise in the amount of attempts the students made once they were given the letter key, but a much greater increase can be noted if only the number of attempts for the first page were tabulated. If (and only if) the study instrument was reduced to a single page, that figure could be seen as an increase of 96%.
These numbers may suggest that the test was, in essense, too long. Once the students were ‘armed with the code-breaking tool’, they seemed to spend more time concentrating on the first page. For many subjects, the allotted 20 minutes ran out before they had time to begin the second or third pages. An additional follow up study, then, would investigate whether pre- and posttests of a single page in length would yield findings of greater signifigance.
B. Change in the Number of Answers Correct (table
|There was a 37% increase in the average number correct, from 37.39 - average pre-test number correct to 51.42 - the average post-test number correct.|
|The pre-test percentage correct was 79% out of the average number attempted (48).|
|The post-test percentage correct was 91% out of the average number attempted (56)|
|The pre-test percentage correct was 22% out of the entire number possible (173).|
|The post-test percentage correct was 30% out of the entire number possible (173).|
In comparing the number of pretest correct answers to the number of answers correct on the posttest, a statistical analysis showed a change signifigant at .5 probability.
Considering the average number correct in relation to the average number attempted rather than to the total (173) possible, there was an increase from 79% to 91%. Viewing the average number correct in relation to the entire number possible the scores show an increase from 22% to 30% average correct.
Although the numbers show only a small change in test scores as the result of using the phoneme relationship key, this researcher would still venture to say that this information can be of considerable pedagogical value. Half (11 of 21) of the subjects volunteered their remarks about this technique for vocabulary learning and all but one of their comments were in favor of the strategy. Six of the subjects qualified their support of the key by saying it needs to be taught or facilitated by an instructor. While this study, in general, leaves doubt as to the effectiveness of using diachronic linguistics in the introductory level German instruction curriculum, it has shown that knowledge of the letter relationships can be helpful and may spark interest in the lexical body of shared German and English lineage. Details of that relationship are part of the historical phenomenon but they also have some significance for present practical use.
Andersen, Roger, Ed. (1984) ‘Second Languages: A Cross-Linguistic Perspective’ Cross- Linguistic Series on Second Language Research. ED255029
Arteaga, Debra L., Herschensohn, Julia. 1995. ‘Using Diachronic Linguistics in the Language Classroom’ The Modern Language Journal. 79-2, 212-225.
Booth, James R., Hall, William S. (1994) ‘Relationship of Reading Comprehension to the Cognitive Internal State Lexicon’ Reading Research Report No. 14.
Bouma, Lowell. 1975. ‘On Teaching the History of German as Applied Linguistics’ Paper presented at a meeting of the American Association of Teachers of German (AATG). Washington, D.C., November 29, 1975.
Braun, Peter. (1985) ‘Internationalisms -- Identical Vocabularies in European Languages’ ED276284
Carroll, Susanne E. (1992) ‘On Cognates’ Second Language Research, 8-2, 93-119.
Clausing, Stephen. 1987. ‘A New Approach to the Use of the Etymological Dictionary in Teaching’ Unterrichtspraxis. 26-2 ,71-75.
Damerau, F. J. (1975) ‘Mechanization of Cognate Recognition in Comparative Linguistics’ Linguistics. 148; 5-29
Dollenmayer, David B. and Hansen, Thomas S. (1984, 1996) Neue Horizonte
DC Heath and Company. Lexington.
Gallegos, Robert L. (1979) ‘Cashing in on Cognates’ Pointer. 23 -3, 10-15
Garrison, David (1990) ‘Inductive Strategies for Teaching Spanish-English Cognates’, Hispania; 73-2, 508-12.
Gass, Susan M., Ed. Selinker, Larry, Ed. (1993) ‘Language Transfer in Language Learning’ Language Acquisition & Language Disorders ED379940
Hancin-Bhatt, Barbara, Nagy, William (1993) ‘Bilingual Students' Developing Understanding of Morphologically Complex Cognates’, Technical Report No. 567.
Holmes, John L. (1986) ‘Snarks, Quarks and Cognates: An Elusive Fundamental Particle in Reading Comprehension’ ESPecialist.15 -13
Martinez, Maria Stella. (1994) ‘Spanish-English Cognates in the Subtechnical Vocabulary Found in Engineering Magazine Texts’, English for Specific Purposes; 13-1, 81-91.
McCrum, Robert., Cran, William., MacNeil, Robert. (1986) The Story of English. Penguin Books. New York.
Nagy, William E. (1992) ‘Cross-Language Transfer of Lexical Knowledge: Bilingual Students' Use of Cognates’, Technical Report No. 558
Nagy,W.,, Garcia. G.E., Durgunoglu, A., & Hancin-Bhatt 1991. ‘Spanish-English bilingual students’ use of cognates in English Reading’ Journal of Reading Behavior. Paper presented at the annual meeting of the American Educational Research Association, Chicago.
Palmberg, Rolf. (1988) ‘Five Experiments of EFL Vocabulary Learning: A Project Report’. Paper presented at the Annual Meeting of International Association of Applied Linguistics. Edinburgh, Scotland.
Pyles, Thomas., Algeo, John. (1982) ‘The Origins and Development of the English Language’ Harcourt Brace Jovanovich. New York.
Spinelli, Emily., Siskin, H. Jay (1987) ‘Activating the Reading Skill through Advanced Organizers’ Canadian Modern Language Review. 44-1, 120-33
Treville, Marie-Claude. (1993) ‘Role des congeneres interlinguaux dans le developpement du vocabulaire receptif: Application au francais langue seconde’ (The Role of Interlingual Cognates in the Development of Receptive Vocabulary: Application to French as a Second language). ED360864
Smith, Sidney. 1968. ‘Historical Linguistics and the Teaching of German’ German Quarterly. 41-2, 231-238.
Appendix A Subjects’ Posttest Remarks
"Cognates could be useful if taught. It would make me feel more comfortable with a foreign language by bring the relation to English more to light."
"The cognat paper was somewhat helpful, but I would prefer actual teaching and learning of the word."
"This would help if taught. Seems to help just by reading the instructions."
"Interesting idea - the rule sheet helped alot - I’m sure there are many exceptions and such. This would be interesting to see if one would learn vocabulary quicker!"
"This would be more effective if lead by a teacher throughout exercise. I could do this by myself, but it would take me much longer."
"Would be a definite help in learning the language."
"Obviously it has to be tough - some are obvious like blood, night and have. Others like cheese need to be taught."
"Knowing the relationships makes it easier to remember the words and what they are. You should give this as homework or an exercise. It helps to learn."
"This is a good idea. It needs to be taught though."
"Its a neat trick, but sounding out the words in my head was as important as the key. I would not like being taught this. I would prefer to just memorize the vocab. I think this would become automatic once enough word that do it were learned."