Looks like SNARC spirit: Coexistence of short- and long-term associations between letters and space

Lilly Roth, Julia F. Huber, Sophia Kronenthaler, Jean‐Philippe van Dijck, Krzysztof Cipora, Martin V. Butz, Hans‐Christoph Nuerk

TL;DR

This study investigates how we associate letters with space, building on existing research about spatial-numerical associations. They found that both long-term knowledge of the alphabet and short-term memory of learned letter sequences influence our spatial responses, suggesting a hybrid model of cognitive representation.

What Problem Does This Paper Address?

For decades, researchers have observed that people tend to associate smaller numbers with their left and larger numbers with their right – the Spatial-Numerical Association of Response Codes (SNARC) effect. The origin of this effect is debated: is it driven by our internal representation of number magnitude (cardinality) or by the ordered nature of numbers (ordinality)? Furthermore, this spatial association has been observed not just for numbers but also for other ordered sequences like letters. This paper aims to disentangle the roles of long-term memory (LTM) for established orders (like the alphabet) and working memory (WM) for newly learned sequences, specifically using letters to isolate ordinality from magnitude.

Methods

The researchers conducted three online experiments using a go/no-go consonant-vowel classification task. Participants were asked to identify letters as either consonants or vowels. The key manipulation across the experiments was the salience of working memory demands.

  • Experiment 1 (Control): This experiment served as a baseline and manipulation check. Participants performed the classification task with letters as "go" stimuli and non-letter symbols as "no-go" stimuli. No specific letter sequence was learned, meaning WM was not actively engaged for sequence memorization.
  • Experiment 2 (Maintenance): Participants first learned a specific four-letter sequence. During the classification task, they were presented with letters from this sequence as "go" stimuli and non-letter symbols as "no-go" stimuli. The learned sequence needed to be maintained in WM for later recall, but not actively retrieved during the task.
  • Experiment 3 (Retrieval): Similar to Experiment 2, participants learned a four-letter sequence. However, in this experiment, letters not in the learned sequence were used as "no-go" stimuli. This forced participants to actively retrieve the learned sequence from WM to correctly identify "go" vs. "no-go" trials.

The researchers analyzed reaction time differences (dRTs) between left and right hand responses, regressing these differences against letter position in the alphabet (for the Spatial-Alphabetical Association of Response Codes, SAARC effect) and position in the learned sequence (for the Ordinal Position Effect, OPE).

Datasets

The study used data from 515 participants recruited online. No new benchmarks or datasets were introduced; the focus was on analyzing behavioral data from the designed experiments.

Key Insights and Ideas

  • Letters have a spatial bias too: Just like numbers, letters show a spatial association, with earlier letters in the alphabet tending to be associated with the left and later letters with the right (SAARC effect). This highlights that ordinality, not just magnitude, can drive spatial associations.
  • WM matters, but retrieval is key: While simply maintaining a learned letter sequence in WM didn't influence spatial responses (Experiment 2), actively retrieving that sequence during the task did (Experiment 3). This suggests that WM's role in spatial associations is most pronounced when the information needs to be actively accessed and used.
  • Coexistence of LTM and WM influences: The study found that both the long-term alphabetical order (SAARC) and the short-term learned sequence order (OPE) can influence spatial responses, and importantly, they can coexist. This supports a hybrid model where both stable, long-term knowledge and dynamic, short-term memory contribute to cognitive processes.
  • Individual letters can have strong biases: The letter 'A', being the first in the alphabet, showed a particularly strong left-hand bias, suggesting that highly overlearned, stimulus-specific associations can sometimes override or influence the effects of learned sequences.

Key Results

  • SAARC effect: A significant SAARC effect (spatial association with alphabetical order) was found in all three experiments, indicating that long-term knowledge of the alphabet consistently influences spatial responses.
  • OPE: An Ordinal Position Effect (OPE – spatial association with the learned sequence) was only detected in Experiment 3, where retrieval of the learned sequence was required during the task. This effect was absent in Experiment 2, where the sequence was merely maintained.
  • Hybrid Model Support: The findings strongly support a hybrid model that integrates both long-term (LTM) and short-term (WM) representations for spatial associations, applicable to both numerical and non-numerical ordinal stimuli like letters.

Limitations and Future Work

  • Sample Size: While larger than many previous studies on OPE, the achieved sample size was smaller than initially planned by the preregistered power analysis. However, the authors note that the sample sizes were still substantial and allowed for significant findings.
  • Online vs. Lab: The study was conducted online, which means controlling for all environmental factors (like noise) was challenging. Participants were screened for noisy environments and data quality.
  • Cultural Bias: The findings are primarily based on participants with left-to-right reading directions and may not generalize to cultures with different writing systems.
  • Stimulus-Specific Effects: The strong bias associated with the letter 'A' suggests that individual stimulus properties can interact with or even override general sequence effects. Future work could explore this further, investigating if other highly overlearned letters (like 'Z') show similar effects and how training studies can illuminate the interplay between LTM and WM associations. The authors also suggest investigating the potential for spatialization to fade over time within WM.

Summary updated

Abstract

Many studies have demonstrated spatial-numerical associations, but the debate about their origin is still ongoing. Some approaches consider cardinality representations in long-term memory, such as a Mental Number Line, while others suggest ordinality representations, for both numerical and non-numerical stimuli, originating in working or long-term memory. To investigate how long-term memory and working memory influence spatial associations and to disentangle the role of cardinality and ordinality, we ran three preregistered online experiments (<i>Ntotal</i> = 515). We assessed spatial response preferences for letters (which only convey ordinal but no cardinal information, in contrast to numbers) in a bimanual go/no-go consonant-vowel classification task. Experiment 1 ("no-go" trials: non-letter symbols) validated our setup. In Experiments 2 and 3, participants learned an ordinal letter sequence prior to the task, which they recalled afterwards. In Experiment 2, this sequence was merely maintained ("no-go" trials: non-letter symbols), whereas in Experiment 3, it needed to be retrieved during the task ("no-go" trials: letters outside the sequence). We replicated letter-space associations based on the alphabet stored in long-term memory (i.e., letters earlier/later in the alphabet associated with left/right, respectively) in all experiments. However, letter-space associations based on the working memory sequence (i.e., letters earlier/later in the sequence associated with left/right, respectively) were only detected in Experiment 3, where retrieval occurred during the task. Spatial short- and long-term associations of letters therefore seem to coexist. These findings support a hybrid model that incorporates both short- and long-term representations, which applies similarly to letters as to numbers.
BibTeX
@inproceedings{roth2025looks,
  author = {Lilly Roth and Julia F. Huber and Sophia Kronenthaler and Jean‐Philippe van Dijck and Krzysztof Cipora and Martin V. Butz and Hans‐Christoph Nuerk},
  title = {{Looks like SNARC spirit: Coexistence of short- and long-term associations between letters and space}},
  booktitle = {Quarterly},
  year = {2025},
  doi = {10.1177/17470218251324437},
}
Figure