vocabulary knowledge on English reading comprehension alsoappeared to be mediated by awareness of cognates.
Perhaps the most important finding of this experiment according to Nagy et al (1992) is the interaction between Spanish vocabulary knowledge andawareness of cognate relationships. Knowledge of Spanish vocabulary can contribute to English readingcomprehension, but this contribution is not automatic. Rather, the contribution of Spanish vocabularyto English reading depends on the extent to which the student is aware of the cognate relationshipsbetween the two languages.
This study supports instruction aimed at increasing Spanish-English bilingual students’ utilization ofSpanish-English cognate knowledge in their English reading. The fact that students as young as fourthgrade demonstrated cognate awareness suggests that explicit instruction in the use of cognates instudents’ second-language reading can help these students overcome difficulties that they may face withEnglish reading vocabulary.
However, there are a number of interrelated factors that may affect bilingual students’ knowledge anduse of cognates in their reading and that may need to be taken into account in the development ofappropriate instructional practices. Degree of bilingualism is one of these factors. All of the studentsin our study were bilingual and biliterate, although most of them were Spanish dominant in theirvocabulary knowledge. The fact that difficult vocabulary in English often has more frequently knowncognates in Spanish should make the transfer of Spanish lexical knowledge to English reading moreadvantageous than the transfer of English lexical knowledge to Spanish reading. Nonetheless, we suspectthat bilingual, biliterate students should be able to use their cognate awareness to help them readunfamiliar vocabulary in either of their two languages. Qualitative findings based on think-aloud withsmaller samples of expert bilingual readers suggest that this, in fact, does occur. Although the extent to which such transfer is beneficial inboth languages still needs further investigation. (Nagy, G, P.9).
2.5. Experiments on English cognate words
Cognate (Latin: cognatusco+gnatus, ie. nasci “to be born”) means: “related by blood, having a common ancestor, or related by an analogous nature, character, or function” (Wikipedia). The clearest cases are those where the parent language is known to exist. For example, on the basis of the various words for father in the Romance languages, such as père and padre, it is possible to work out how they all derived from the Latin word pater. If Latin no longer existed, it would be possible to reconstruct a great deal of its form, by comparing large numbers of words in this way. A word that is related in origin to another word, such as English brother and German Bruder. In linguistics, cognates are words in one or more languages that have a common origin, meaning that they are descendents of a same word, possibly in a common predecessor language. One example is the English word night, whose cognates include words with the same meaning in other languages, like nuit (French), Nacht (German), and nakti- (Sanskrit). In less formal usage, “cognate” may simply mean a synonymous relationship between words.
Cognates may also arise through borrowings into languages.the word tangerine in English or spinach originates to their Persian base form “نارنگي” and “اسفناج”.
According to Talebnejad et al (2012) the similarity of words between languages is not enough to demonstrate that the words are related to each other, in much the same way that facial resemblance does not determine whether two people are genetically related. Over the course of hundreds and thousands of years, words may change their sound completely. Thus, for example, English five and Sanskrit pança are cognates, while English over and Hebrew a’var are not, and neither are English dog and Mbabaram dog.
The history of English cognate words
According to Harris (2001), most of the ancient and modern languages of Europe belong to a family of languages which is called by modern scholars “Indo-European” and their study falls within the range of research known as Historical Linguistics. It was first noticed by Sir William Jones, a linguistically minded employee of the British East India Company in the late 18th century as he began private lessons in Sanskrit, that most of the languages of Europe bore a strong resemblances to each other in basic, primary vocabulary. These languages furthermore seemed connected structurally with the ancient Sanskrit which he was learning.
Perhaps about 10000 B.C, when successive waves of peoples poured out of an original area north of the Black sea and west of the Caspian. One group went south into Persia and India, some went West across Europe, closely following each other and producing what would later become the Greek, Italic (including Latin), Celtic, Germanic, and to the north, Baltic and Slavic speaking populations. All of these languages are clearly related, but their relationships can be understood only in the light of a web of complex sound laws, which constitute the present discipline of Indo-European Historical Linguistics. (Harris, W, 2001).
Words are the elements of a language used to identify objects, ideas and express feelings. These words are also the mirrors in that they are a reflection of the cultural history and development of a people. Tracing words to their origins will open a new window to human civilization and culture and help us understand the roots of some of our present social trends and attitudes. It is likely to consider as vital characteristic for teachers to be aware of the misuse of learners to enhance communication and improve language learning, and find convenient solutions and recognize those misbehaviors to get efficient result. This study investigates that knowing the origin of learners ‘misbehaviors (as a result of using false friends) eradicate miscommunication in the process of language learning. Its effect should be examined based on learners ‘reflections. False friends have been widely considered in different language areas: translation studies, language teaching, lexicography or contrastive linguistics. (Talebinejad et al, 2012, P.1477).
One of the variables that affect the way words are represented in abilingual memory is the characteristics of the word. One of thesecharacteristics examined in a variety of studies and across a variety oflanguages is the “cognate status” of the translation pair. Some empiricalstudies focused on cognate and noncognate difference (Gollan et al.,1997;Lalor, &Kirsner, 2001). Noncognates are translation equivalentswith different spellings and sound patterns in the two languages (e.g., thePersian word /sabz/ and its English translation green), whereas cognatesare translation equivalents with similar orthographic or phonetic form. (Kondrak, Marcu, & Knight, 2003). The similarity is usually due to eitherhistorical reasons (e.g., the Persian word /lab/and its English translationlip) or borrowing from one language to another (e.g., the Persian word/keyk/and its English translation cake).
Bilingualsstrategically connect one language with the other by detecting therelationship between the prime and the target stimulus (Kirsner et al,1984. (A way to hide the bilingual nature of the task is using masked priming paradigm developed in the studies of visual word recognition(Evett& Humphreys, 1981; Forster & Davis, 1984). In this paradigm, avery briefly presented prime preceded by a forward mask (like a numberof signs) is immediately followed by a given target stimulus so that theprime cannot be identified. Due to the adopted masking procedure, theprime is, for most participants, virtually invisible and cannot beidentified.
Using the masked priming paradigm, some empirical studies havecompared the effect of priming for cognates with noncognates (Altarriba,1992; Chen & Ng, 1989; Cristoffanini, Kirsner, &Milech, 1986; deGroot&Nas, 1991; Goll
an et al., 1997; Jin, 1990; Keatley& de Gelder,1992; Williams, 1994). Essentially, these studies have investigatedwhether words sharing semantic, orthographical, and phonologicalrepresentations (cognates) are processed differently from those sharingonly semantic representations (noncognates) under the masked primingparadigm.
Cognate words have shown easier and faster processing thannoncognates in isolated word recognition tasks such as lexical decision(Dijkstra, Grainger, & van Heuven, 1999; Lemhofer&Dijkstra, 2004;Lemhofer, Dijkstra, & Michel, 2004). This task is a procedure thatmeasures the magnitude of translation priming across differentlanguages; it involves measuring how quickly participants classifystimuli as words or nonwords. This task is mostly used for identifyingprocessing similarities and differences between the main and the controlgroups.
On the whole, cognate translations have been shown to be processedfaster in a number of experiments as compared with noncognatetranslations. However, the results concerning noncognates is somewhatmixed. On the basis of these results, de Groot and Nas (1991) andSanchez-Casas et al. (1992) suggested that cognate translations mayshare common representations in memory, whereas noncognatetranslation equivalents do not. However, de Groot (1992) changed herposition regarding this issue later and suggested that noncognates maysimply share fewer nodes at the conceptual level than do cognates.
Assuming that cognates share the same representations in memory, a number of studies focused on the role of orthography in establishingshared lexical entries for cognates in a bilingual memory. Theyinvestigated whether both orthographic and phonological overlaps arerequired for establishing such entries or whether orthography has no rolein this process. In an attempt to test languages with different scripts,Bowers, Mimouni, and Arguin (2000) failed to find any priming forArabic-English, whereas a significant priming was obtained fororthographically similar languages (e.g. de Groot &Nas, 1991; Sanchez-Casas et al, 1992). Therefore, it can be concluded that orthography playsa role in obtaining long-lag cognate priming effects. In another study byGollan et al. (1997), four experiments were designed to examine thenecessity of orthographical overlap in obtaining significant cognateeffects. Both cognates and noncognates were tested in the experimentswith lexical decision task for the purpose of comparison. The results ofthe study showed that, in contrast with Bowers et al.’s (2000) study, enhanced cognate priming was observed despite the absence oforthographical overlap. One noticeable point about this study was that, unlike previous studies, cognate priming was found only with L1 (i.e.,the dominant language) but not with L2 (i.e., the nondominant language) primes. The results of the study were interpreted in terms of a duallexicon model according to which both orthographical and phonologicaloverlaps are needed to establish shared lexical entries for cognates. Asthere are only few studies testing cognates across languages withdifferent scripts in both directions, it is not totally clear whether thecognate effect is purely phonological or the joint effect of bothorthographical and phonological similarities. Therefore, it is important toinvestigate whether the predictions of dual lexicon model regarding thenecessity of orthographic overlap for establishing shared lexical entriesfor cognates in a bilingual memory can be confirmed in more studies.
According to Pirooz (2004, P.119), Due to their nature, cognate pair may be either true cognates or false cognates. “True cognates” are found when these related words have more or less the same meanings as their forms may suggest. That is, their spelling, pronunciation, or both may suggest similar senses in the two cognate languages. For example, maadar and mother in