Recognition without awareness may seem improbable, yet it offers intriguing insights into cognitive processes. This phenomenon suggests that recognition does not always require conscious and active engagement; it can occur unconsciously. The authors of this study aimed to investigate whether recognition can transpire without conscious awareness through a series of four-alternative forced-choice (4-AFC) recognition tests. 

The AFC recognition test is a quantitative and robust assessment of recognition memory. Participants were first presented with a series of stimuli to study, called the encoding phase, followed by a series of trials. Each trial involved the presentation of either 2 (in a 2-AFC test) or 4 (in a 4-AFC test) stimuli, one of which was from the original list. Participants were then asked to select the stimulus from the original list. This design ensured that participants could not be ambiguous about their choices without a basis from their memory, thereby minimising the potential for biases and guesses. 

The authors conducted four experiments, each with a similar methodology. During the encoding and testing phases, participants were randomly assigned to either a full attention (FA) group or a divided attention (DA) group. Including DA conditions was crucial to the study as attention during recognition memory tests can enhance performance. Given that the primary goal of this study was recognition without awareness, the comparison between FA and DA conditions helps the authors understand the role of awareness, or the lack thereof, in encoding and recognition memory, thereby underscoring the importance of this research. 

Before the 4-AFC tests, participants were informed that some trials might not include any stimuli from the original list, though they were not told which trials these would be. Participants were instructed to choose an option even if they did not recognise any targets from the original list. To gauge confidence, participants rated their choices on a scale from 0 to 2: a rating of 2 indicated high confidence that the item was on the list, 1 suggested possible recognition, and 0 signified a forced guess. Recognition without awareness was inferred when the target items from the original list were present during the testing phase, and the participants rated their choices as 0 yet chose correctly at a p-value above 0.25, i.e., above chance.

In the first experiment, participants initiated the encoding phase, followed by a 10-minute retention period where they played a game of Tetris and concluded with the 4-AFC test. The encoding phase involved presenting participants with 40 common nouns in various font colours (red, blue, green, and yellow) and requiring them to judge the font colour quickly while ignoring the noun. The DA group in this experiment listened to a string of odd numbers as a distraction. Results following the testing phase indicated that participants in both conditions who rated their choices as 0 but guessed correctly did so significantly above chance levels. 

To substantiate their findings, the second experiment modified one aspect: participants were informed that their colour-judging performance during the encoding phase would be tested later. The authors hypothesised that this would increase deliberate encoding and decrease the frequency of 0 confidence ratings. This change would also help determine whether recognition without awareness occurred due to the incidental encoding phase, as the participants were unaware their encoding would be tested later (as in the first experiment). Following the experimental phase, they again found results in favour of recognition without awareness, suggesting the phenomenon would occur regardless of whether the encoding phase was incidental or intentional. The number of 0-rated choices decreased, but the hit rate increased only in FA conditions.

Contrary to the authors' predictions, the proportion of correct answers rated 0 increased in FA conditions and only insignificantly decreased in DA conditions. Despite suboptimal results for recognition without awareness under DA conditions in the first two experiments, the authors proceeded with a third experiment, altering the encoding paradigm. Instead of a colour-word association, participants judged word pairs during the 4-AFC test. Although FA conditions remained unchanged, DA conditions were slightly less distracting. Nonetheless, this experiment also failed to yield significant results in favour of recognition without awareness under DA conditions. 

The authors recognised the importance of context regarding recognition and memory, which is why the final experiment in this study involved a test for context change. They achieved this by replicating experiment three with one minor tweak: every 4-word set during the 4-AFC test would be preceded by a word from the original list studied during the encoding phase. For example, if one of the word pairs studied in the encoding test were MOTH-FOOD, the word "MOTH" would precede the word set (BASE, FOOD, BOOK, FARM) during the recognition test. However, since the test was designed so that only half of the word sets during the recognition test contained a target word from the original list, in trials where a target word was not present, a random word from the original list would precede the word sets. This manipulation allowed the authors to predict an increase in hit rate, a reduction in choices rated 0 and, thereby, a significant decline in the phenomenon of recognition without awareness. Upon concluding the experiment, their results suggested a predicted recognition performance increase. However, adding context during the testing phase did not significantly impact the recognition phenomenon without awareness, nor did it suggest a favourable result for DA conditions. 

Across the four experiments conducted by authors, they consistently found results for recognition without awareness, suggesting implicit influences outside of one's conscious attention may gravitate them towards the correct choice, even though they may consider them pure guesses. Participants performed above chance in all experiments, although the authors could not conclusively determine if this phenomenon was more prevalent under DA conditions. They also concluded that general task difficulty could influence recognition; easier tasks led participants to be more conservative in their guesses, while more challenging tasks resulted in fewer 0 ratings, indicating participants felt more certain of their choices.

​​

​

Citation: 

​​

Craik, F. I. M., Rose, N. S., & Gopie, N. (2015). Recognition without awareness: Encoding and retrieval factors. Journal of Experimental Psychology: Learning, Memory, and Cognition, 41(5), 1271–1281. http://dx.doi.org/10.1037/xlm0000137