A Psycholinguistic Investigation of Emotion Words in Turkish
Filiz MergenEmotion words, which are verbal expressions of emotions, have long been the subject of psycholinguistic studies. It was revealed that emotion words are different from neutral words and need to be classified accordingly, even within themselves. To put differently, it has been accepted that there are differences between emotion words that directly refer to emotions and emotion-laden words that are associated with emotions. In this study, it is aimed to reveal if Turkish emotion words differ based on this criterion. Native speakers of Turkish (N=39) were asked to distinguish these two types of words from letter strings which are potentially legitimate words, but lack conceptual associations (non-words). They were asked to decide whether the stimuli presented on a computer screen were real words or not. Their reaction times and accuracy of responses were recorded. The results revealed that there was a temporal difference between the words in both categories. While emotion words were processed faster than emotion-laden words. In addition, it was shown that all words that indicate or elicit positive emotions, regardless of the type of word, were processed faster than negative words. The same result was also obtained in distinguishing real words from non-words. These findings were discussed with regard to psycholinguistic principles.
Türkçedeki Duygu Sözcüklerinin Psikodilbilimsel Açıdan İncelenmesi
Filiz MergenPsikodilbilim alanında, duyguların dil ile ifade edilmesini sağlayan duygu sözcükleri uzun süredir birçok çalışmanın konusu olmuştur. Duygulara işaret eden sözcüklerinin zihinsel işleyişini ele alan çalışmalarda bu tür sözcüklerin diğer sözcüklerden farklı olduğu, hatta kendi içlerinde bile sınıflandırılması gerekliliği ortaya çıkmıştır. Buna göre, doğrudan duygulara işaret eden duygu sözcükleri ile, duyguları çağrıştıran duygu yüklü sözcüklerin arasında farklılıkların olduğu kabul edilmiştir. Bu farklılıkların Türkçede nasıl olduğunu hedefleyen bu çalışmada, ana dili Türkçe olan bireylerden bu iki tür sözcüğünü anlamı olmayan harf dizilerinden ayırt etmeleri istenmiştir. Bilgisayar ekranından sunulan bu uyaranların gerçek sözcük olup olmadığına karar vermeleri istenmiş ve bu görev esnasında tepki süreleri ve cevaplarının doğruluğu kaydedilmiştir. Elde edilen bulgulara göre, her iki kategoride bulunan sözcüklerin arasında zamansal olarak farklılık olduğu ortaya çıkmıştır. Duygu sözcükleri daha önce farkedilirken, duygu yüklü sözcüklerin daha uzun sürede işlemlendiği görülmüştür. Ayrıca, sözcük türüne bağlı olmaksızın olumlu duyguları uyandıran ya da işaret eden tüm sözcüklerin olumsuz yüklü sözcüklerden daha hızlı işlemlendiği ortaya çıkmıştır. Aynı bulgu, gerçek sözcüklerin anlamsız harf dizilerinden ayırt edilmesinde de görülmüştür. Bu bulgular psikodilbilimsel prensipler çerçevesinde değerlendirilmiştir.
Introduction
Emotions are critical in our lives, and language is undoubtedly the most sophisticated means to convey them. Psycholinguistic studies investigating word recognition have long underestimated the role of emotion content of words, and considered emotion words as comparable to neutral words. Following Morton (1979, cited in Jared, Poh & Paivio, 2013), who suggested that concrete words have both verbal and visual associations while abstract words can only be labelled verbally, this line of research, then, categorized emotion words as abstract words. However, emotions’ distinctive role have prompted researchers into examining them in verbal communication. It was Altarriba, Bauer and Benvenuto (1999) who first suggested that emotion words form a distinct category and should be classified separately. Later research both confirmed and strengthened this view, leading to the distinction even among emotion word category. Accordingly, it was argued that words in this category were different in terms of their conceptual associations such that some words directly refer to affective states such as sadness, happiness or anger (emotion words), while others (emotion-laden words) relate to emotions indirectly (Altarriba, 2006; Pavlenko, 2008). To put differently, the former labels emotions while the latter elicits emotions. Based on the idea of either direct or indirect reference to emotional states, it is argued in the literature that emotion words are more easily recognized than emotion-laden words. This is called Mediated Account, according to which, in the recognition of emotion-laden words, conceptual access comes first and is followed by access to affective meanings, leading to longer and more effortful processing, while in the recognition of emotion words, affective meanings are directly accessible, enabling faster recognition. In addition to the role of word types in the psycholinguistic aspects of word recognition literature, lexical features like valence (i.e. whether a word is positive or negative) and word advantage (i.e. whether a word is meaningful or not) have been a topic of interest. Extant literature investigating the former has firmly established that positive words have a processing advantage compared to negative words (Positivity Effect). Evolutionary perspective relates this priviledge to our tendency to approach to positive stimuli and ward off negative stimuli to ensure the survival of the species, as the latter is taken as a threat to survival (Taylor, 1991; Estes & Adelman, 2008). Similarly, Word Superiority Effect is known as the ease and comfort with which one can discriminate meaningful words from non-words. This has been attributed to the availability of conceptual associations of real words, which non-words lack although they can be phonologically and orthographically legitimate words in a given language.
Method
The main focus of this study is to investigate the difference between the two types of words in Turkish. Following the literature, it was also aimed to disclose whether the two widely accepted phenomena, namely Positivity Effect and Word Superority Effect, will be replicated in Turkish. To this end, thirty-nine participants were recruited to take part in a lexical decision task in which they were told to decide whether the word strings they were presented on the screen were real words or non-words. Based on the native speakers’ categorization of Turkish words from a pool of 180 words, the stimuli used in the study consisted of a total of forty words categorized as emotion and emotion-laden words (20 positive and 20 negative). Also, forty non-words were formed by changing the letters of real words in Turkish. Following a trial session to familiarize the participants with the procedure, the stimuli were presented in two blocks. The data was analyzed by using 2 (emotion vs. emotion-laden words) x 2 (positive vs. negative) ANOVA for Repeated Measures.
Results and Discussion
The results were in line with those obtained in the literature. First, it was found that emotion words were processed significantly faster than emotion-laden words. This result can be explained by Mediated Account, which predicts longer response times due to indirect access to emotion-laden words. This view has found considerable support in the literature (Altarriba & Basnight-Brown, 2011; Kazanas & Altarriba, 2015, 2016; Liu et al., 2022: Wu et al., 2021a). Second, in line with the hypothesis of the current study and the relevant literature (Kousta, Vinson & Vigliocco, 2011), positive words in Turkish were processed faster than negative words. Considering our evolutionary background, the privileged status of positive emotions is an expected outcome. Finally, Word Superiority Effect was investigated, and it was found that real words were processed faster. These results confirm and extend the current results in the psycholinguistic literature and provide support from Turkish language.
