The Frequency of Sensory Adjectives and the Juxtaposition of Nouns and Sensory Adjectives in Middle Persian Zoroastrian Texts From a Sensory Linguistics Perspective

This study is a descriptive–analytical and data-based research. From the perspective of sensory linguistics, the frequency of sensory adjectives and the juxtaposition of nouns and sensory adjectives were examined in three Middle Persian Zoroastrian texts. The aim was to determine the extent to which each of the five senses was represented by sensory adjectives in the texts and whether the distribution corresponded to the hierarchy proposed by Lynott and Connell (2009) from highest to lowest. In another part of the research, after analyzing the nouns and juxtapositional sensory adjectives, it was investigated whether the nouns and adjectives within the same sensory domain exhibited the highest frequency of sensory juxtaposition or not. The findings showed that the largest number of sensory adjectives was related to the visual domain, while the smallest number was associated with the sense of smell, which is consistent with the hierarchy of the senses suggested by Lynott and Connell (2009) and Winter (2019). Moreover, the analysis of the frequency of nouns and sensory adjectives revealed that, except for the sense of sight, nouns and adjectives of the same domain did not occur more frequently than those in different domains. Therefore, this finding does not fully align with the claims of Lynott and Connell (2009) and Winter (2019). 
Introduction
The senses allow humans to interact with and perceive their environment, forming a crucial part of our being. Their number is debated due to overlap and complexity. Aristotle categorized five basic ones, each linked to specific organs: vision with the eyes, hearing with the ears, touch with the skin, smell with the nose, and taste with the tongue (and nose). Neuroscientists identify at least nine senses, including Aristotle's five senses plus pain, pressure, body awareness, and temperature. Some replace pressure with balance or add balance to the list. Psychologists define even more senses, considering both nerves and the mind.
Since each sense is represented unevenly in language, this study, within the framework of sensory linguistics, investigates the frequency of each sense in sensory adjectives in three Middle Persian Zoroastrian texts (the Bundahišn, the Vizidagīhā ī Zādisparam, and the Ardā Wirāz Nāmag). It also examines the juxtaposition of nouns and sensory adjectives to determine whether same-domain pairings are more frequent than cross-domain ones. This study addresses two questions:

What proportion of sensory adjectives in the three Middle Persian Zoroastrian texts is associated with each of the five senses, and does this distribution align with the hierarchy of these senses proposed by Lynott and Connell (2009)?
Are sensory nouns and adjectives from the same domain more frequent in juxtapositions in these texts, and if so, why?

This research follows the theoretical framework of Lynott and Connell (2009) and Winter (2019). Lynott and Connell presented 423 sensory adjectives to participants, asking them to associate the adjectives with one or more senses. Their study revealed that most adjectives were linked to multiple senses. They aimed to determine which sense was associated with the highest number of sensory adjectives in English—that is, how the average perceptual strength of the senses is distributed in English, and which sense has the most and the least perceptual strength. Using a five-point Likert scale, they presented 423 English sensory adjectives to 55 native English speakers and found the following average perceptual strength of the senses for these adjectives:        
Vision > Touch > Hearing > Taste > Smell (Lynott & Connell, 2009, p. 562)
In the next phase, these researchers asked participants to select a noun for each of the 423 sensory adjectives to examine whether they would choose nouns from the same sensory domain as the adjectives. They hypothesized that same-domain noun-adjective pairs would be more frequent than cross-domain ones.
Method
This study is descriptive–analytical and data-based. The data came from three Middle Persian Zoroastrian texts: the Bundahišn (Pakzad, 2005), the Vizidagīhā ī Zādisparam (Rashid-Mohassel, 2006), and the Ardā Wirāz Nāmag (Gignoux, 2003). In this research, all 240 sensory adjectives in these texts—including simple, derived, and compound forms—were examined. The study was conducted in two stages. First, all sensory adjectives, whether accompanied by a noun or used independently, were listed and categorized into one of the five sensory domains. In this phase, each adjective was counted only once; in other words, all three texts were treated as one corpus. In the second phase, the sensory adjectives used with nouns were analyzed to determine whether the noun and adjective belonged to the same or different sensory domains. When a noun had multiple adjectives, it was counted according to the number of adjectives.
Results
Based on the study’s findings, of the 240 adjectives, vision accounted for the most sensory adjectives (192). Touch appeared significantly less frequently (26). The other senses were closer in frequency (taste 10, hearing 7, and smell 5) and far fewer compared to vision. Overall, the frequency of sensory adjectives in the three Middle Persian Zoroastrian texts can be represented as follows:
Vision > Touch > Taste > Hearing > Smell
Regarding vision, same-domain juxtapositions (154) were more frequent than cross-domain ones, supporting Lynott and Connell’s (2009) and Winter’s (2019) hypothesis. In hearing, same-domain juxtapositions (0) were fewer than cross-domain ones (3), contradicting the hypothesis. Smell did not appear as a source domain, so only cross-domain cases were considered. For touch, cross-domain juxtapositions (4) outnumbered same-domain ones (2) and for taste, cross-domain juxtapositions outnumbered same-domain ones (12 to 3). Therefore, except for vision, findings do not support Lynott and Connell’s (2009) and Winter’s (2019) hypothesis. Thus, same-domain juxtapositions were more frequent only for vision, and this pattern did not hold for the other senses. It is also important that the overall number of data points for the other senses, particularly hearing and smell, was limited in the three texts examined.
Conclusion
According to the results of this study, the senses with the highest and lowest frequency in the three Middle Persian Zoroastrian texts, and thus the highest and lowest average perceptual strength, were vision and smell, respectively. This finding aligns with Lynott and Connell (2009) and supports Sharifi and Sabouri (in press), who argue that the extremes of the sensory hierarchy tend to remain stable across languages and historical stages, while differences appear in the middle senses. In this study, compared to Lynott and Connell (2009), the only difference was in the order of hearing and taste, which was reversed in the Middel Persian Zoroastrian data. However, the number of data points in this study is not sufficient to deem this difference significant.
Furthermore, a comparison of the present study’s findings with Sharifi and Sabouri (in press) regarding the average perceptual strength of senses in modern Persian reveals no difference between Middle Persian Zoroastrian and modern Persian in the average perceptual strength of the five senses. In other words, from Middle Persian to modern Persian, there has been no change in the average perceptual strength of the five senses.
Regarding the hypothesis that same-domain noun-adjective juxtapositions are more frequent than cross-domain ones, this study found that, except for vision, it did not hold for other senses. Comparison with Sharifi and Sabouri (in press) reveals that instances supporting Lynott and Connell’s (2009) hypothesis in Middle Persian Zoroastrian (one instance for vision) were fewer than in modern Persian (three instances for vision, hearing, and taste). Nonetheless, neither study fully supports Lynott and Connell’s (2009) and Winter’s (2019) hypothesis.
Ethical Considerations
Not applicable
Funding
Not applicable
Conflict of interest
The authors declare no conflict of interest.