Is there a graphic accent among Latin-Arabic biscriptual writers?

Technical Summary: Graphic Accent in Latin-Arabic Biscriptual Writers

Problem Statement
Handwriting is a complex fine motor skill shaped by both intrinsic biomechanical constraints and the cultural demands of specific writing systems. While bilingualism in spoken language is known to produce accents, the existence of a "graphic accent" in biscriptual writers—individuals proficient in two distinct writing systems—is less understood. Specifically, it remains unclear whether mastering two scripts with opposing directionalities (Latin: left-to-right; Arabic: right-to-left) and distinct letter formation rules leads to cross-script interference. This interference could manifest as a transfer of motor features, altering spatial dimensions, kinematic fluency, or pen orientation in one script due to the influence of the other. The study investigates whether Lebanese Latin-Arabic biscriptuals exhibit such a graphic accent compared to monoscriptual Latin (French) and monoscriptual Arabic (Syrian) writers, and whether this accent is perceptible to observers.

Methodology
The research employed a two-experiment design involving kinematic analysis and visual categorization.

• Experiment 1 (Production):

  • Participants: Three groups of right-handed adults: 21 Latin-Arabic biscriptuals (Lebanese), 39 Latin monoscriptuals (French), and 22 Arabic monoscriptuals (Syrian).
  • Task: Participants copied five specific words in both French and Arabic using an ink pen on a graphic tablet (Wacom Intuos Pro, 200 Hz sampling).
  • Measures: Spatial variables (word width, height, trace length), kinematic variables (mean velocity, dysfluency, number and duration of pen lifts), and pen orientation (altitude and azimuth angles).
  • Analysis: Linear Mixed Models compared biscriptuals against monoscriptuals within each writing system. Supplementary analyses examined pen orientation differences between scripts and groups.

• Experiment 2 (Perception):

  • Participants: 20 Latin monoscriptuals and 20 Latin-Arabic biscriptuals (new sample).
  • Task: Participants visually categorized anonymized handwritten sheets (produced in Experiment 1) as either "monoscriptual" or "biscriptual."
  • Analysis: A two-way nonparametric ANOVA (Aligned Rank Transform) and Wilcoxon signed-rank tests compared categorization accuracy against chance levels (50%).

Key Contributions and Results

1. Spatial Interference (The "Graphic Accent"):

   • Latin System: Biscriptuals exhibited horizontal expansion (wider words) compared to Latin monoscriptuals, though trace length remained unchanged. This is hypothesized to be an interference effect from Arabic, where kashida (letter elongation) increases horizontal extent.
   • Arabic System: Biscriptuals exhibited vertical expansion (taller words) compared to Arabic monoscriptuals, without increasing trace length. This is attributed to the transfer of Latin features, such as ascenders and descenders.

2. Kinematic and Strategic Differences:

   • Speed: Biscriptuals wrote significantly faster in Arabic than Arabic monoscriptuals, potentially facilitated by their vertical expansion.
   • Pen Lifts: Biscriptuals performed fewer pen lifts in the Latin system than Latin monoscriptuals. Further analysis revealed this was due to a distinct strategy for diacritics: biscriptuals tended to write the entire word body first and then add accents/dots, whereas monoscriptuals often placed accents immediately after individual letters. This strategy mirrors the Arabic writing convention where diacritics are added after the main word structure.
   • Pen Orientation: Biscriptuals adjusted their pen azimuth (orientation relative to the writing surface) based on the script. When writing Arabic, their orientation matched Arabic monoscriptuals. When writing Latin, their orientation was intermediate between the two monoscriptual groups but closer to the Arabic style, suggesting a compromise strategy to minimize postural constraints between opposing writing directions.

3. Perceptibility of the Accent:

   • Monoscriptual Observers: Performed at chance level (approx. 50%) when trying to identify biscriptual writing, indicating the graphic accent is subtle and not visually obvious to those unfamiliar with the cross-script interference.
   • Biscriptual Observers: Significantly outperformed chance (60.3% accuracy) in identifying biscriptual writing. This suggests that biscriptual readers possess a specific perceptual or motor familiarity with the subtle characteristics of biscriptual handwriting.

Significance and Claims
The paper claims that the coexistence of two writing systems influences handwriting production, resulting in a detectable "double graphic accent." This accent is characterized by:

• Spatial Transfer: Horizontal expansion in Latin and vertical expansion in Arabic.
• Motor Strategy Transfer: The application of Arabic diacritic placement strategies (writing the word body first) to Latin script.
• Adaptive Orientation: A unique pen orientation in Latin that serves as a compromise between the two systems.

The authors argue that these findings support the hypothesis that handwriting relies on shared motor-control mechanisms susceptible to interference. Furthermore, the study suggests that the perception of these subtle motor signatures is grounded in sensorimotor experience; only biscriptual observers, who share the motor history of the writers, can reliably detect the accent. This aligns with embodied cognition theories, positing that the sensorimotor representations acquired through writing practice contribute to both production and the perception of written traces.

The authors remain modest regarding the universality of these findings, noting that the study focused on right-handed Lebanese writers and that the accent might be influenced by specific educational contexts (e.g., French-influenced curricula in Lebanon). They conclude that while the accent is a persistent feature of adult biscriptual writing, it is subtle enough to escape monoscriptual detection, highlighting the complexity of motor program interaction in biscriptuality.