Everything, Everywhere…All At Once?

by Phoebe Matthew

art by Luciana Piro

“I have too much to do and not enough time.” This is a familiar dilemma, and multitasking may seem like the only solution; checking off two, three, or more tasks with the time commitment of one is a deal that not many people would pass up. Whether it's listening to music while you drive, replying to emails during a lecture, or calling a friend while you do your homework, we have all normalized the idea of multitasking in our daily lives.  Multitasking appears as a straightforward and universally applicable approach. The neural mechanisms, however, are much more complicated than this easy fix implies.

Bottleneck Model

Multitasking makes people feel more productive and efficient, but can the brain truly process stimuli and command actions simultaneously? A theory that emerged to answer this question is the bottleneck model. According to this model, certain mental operations require exclusive access to central processing resources, meaning the brain can only perform one of these operations optimally at a time [1]. This theory was established by studies using the psychological refractory period (PRP) paradigm, which is defined as an experiment where participants are presented with two different stimuli in succession, each associated with a respective task. [2, 3]. An example of a PRP experiment would involve presenting participants with an auditory stimulus, which they must indicate as low- pitched or high-pitched by pressing a corresponding button in a set, and a visual stimulus of a letter, which they must indicate as an “H” or an “S” by pressing another set of buttons. It has been observed that when the time between the presentation of the two stimuli, known as stimulus onset asynchrony, increases, participants respond faster to the second task [3]. This means that when tasks are spaced farther apart, execution efficiency increases.

The decrease in efficiency when stimuli are presented closely together can be partly explained by the demands placed on the executive function of working memory, which is how the brain stores information and tells you what to use it for [4]. Experiments have shown that working memory is impaired while multitasking. According to the bottleneck model, multitasking involves task switching at one stage of task completion [2]. During this process, information about the other task must be held in working memory [4]. If you are writing an essay while calling a friend, you may lose your train of thought for the essay after responding to a question from your friend. This indicates that multitasking creates more opportunities for errors because it increases dependence on working memory, which may not always be accurate.

To truly understand the neural processes involved in multitasking, it is important to identify the bottleneck, or, in other words, the mental operation that requires exclusive access to the brain's central processing resources. Each task can be simplified into three main stages: perception, response selection, and response production [2, 5]. 

These task stages are carried out in day-to-day life, for example, when you are driving towards a traffic light.  Seeing the light turn red is perception, deciding to move your foot over to the brake is response selection, and actually pressing the brake is response production. Several experiments have been conducted to determine the stage at which the bottleneck occurs [6–8]. No significant reaction time changes are observed when the intensity (high vs. low contrast) of the second stimulus, a visual stimulus, is altered. This indicates that these perceptual differences do not ultimately influence task completion. When multitasking, decreased efficiency is not due to our brain being unable to perceive two stimuli simultaneously. While driving a car, you would be able to hear your phone telling you to make a turn and see the light turn red at the same time. Therefore, the possibility of a perception bottleneck can be ruled out [8].

  A response production bottleneck has not been supported either. In our traffic light example, pressing three pedals in succession to stop the car, rather than just one, would not influence the reaction time for a second task, such as adjusting the volume of the music you are playing. Since increasing the time required to produce the response for the first task doesn’t significantly affect the reaction time for the second task, the bottleneck does not occur in the response production stages [7]. Instead, evidence suggests that the bottleneck occurs at the response selection stage. When additional choices were added as possible responses to the first task, reaction time decreased for both simultaneously presented tasks [6]. For example, you would need to press a distinct pedal when you see a red, yellow, or green light rather than just having 2 options. You would have a slower reaction to pressing the pedals while simultaneously turning the volume of the music down because of the slowed response selection stage for the first task. Therefore, while simultaneous processing is possible at the perception and response production stages, the bottleneck effect can be observed at the response selection stage, which requires exclusive access to central processing systems.

Central Resource Sharing However, central processing may be more flexible than the strict bottleneck model assumes [3]. The response selection bottleneck model imposes an all-or-nothing structural limitation, requiring response selection to occur step-by-step, whereas the central resource sharing theory suggests that the central processing must be split across tasks and processed simultaneously [5, 9]. The central resource sharing theory also assumes that individuals can flexibly allocate their limited resources to different tasks in a proactive (before the stimuli are presented) or reactive (after the stimuli are presented) manner [3]. For example, when listening to music while studying, you may proactively decide to focus almost entirely on your study materials, but when your favorite song comes on, you may react by getting distracted from your work. It may be expected that as one task becomes more predictable, requiring fewer cognitive resources, the remaining resources can be reallocated to the second task. In the context of multitasking, cognitive resources involve working memory capacity and attentional control [4, 10, 11]. Rather than attention and working memory being equally divided among two tasks, our proactive and reactive allocation can divide them more efficiently. However, the idea of automatic reallocation based on task predictability is contradicted by experiments resulting in improvements only in the predictable task [12]. Improvement of reaction time for both tasks only occurred when the structure of one task predicted the outcome of the other task. This association between the stimuli for both tasks caused participants to mentally integrate the two tasks, therefore increasing their simultaneous processing efficiency [12]. The strict divide between working memory and attention for both tasks is broken down by integration, allowing resource allocation to become more efficient. This experiment reveals that it is not enough for one task to become predictable for resource reallocation to occur because the mental separation between the two tasks still exists. It is only when the two tasks become linked or dependent on one another that increased efficiency in processing is enabled. 

Often, central resource sharing theory and bottleneck models make similar predictions for various PRP paradigm experiments, but small distinctions exist. One of these distinct predictions involves the performance of the second task when done separately from the first task. According to the bottleneck model, the first task would be performed equally, if not more efficiently, than the second one, because separate tasks would have total access to the response selection bottleneck [3, 5, 9]. Proactive allocation in the central resource sharing theory predicts that participants might not allocate all resources to the first task in

anticipation of the second task [9].

This can be illustrated by trying to read a book while you're on hold on the phone. You won’t be focused on the elevator music that’s playing on loop, but you may no longer devote your full attention to reading, so that you don’t miss the person when they come back on the line. After the phone call is over, the book you are reading receives your undivided attention. The phone call never truly had all your mental capacity allocated towards it because your mind would have been dwelling on your book. Upon completion of the first task, all resources can be allocated to the second task, so the central resource-sharing model predicts that when the stimuli are presented further apart, the first task is completed less efficiently than the second.

This prediction directly opposes the response selection bottleneck model, but is consistent with experimental results supporting the central resource sharing model [3]. While both models predict decreased efficiency in task completion when stimuli overlap, the bottleneck model suggests strict sequential processing at the response selection stage, while the central resource sharing model allows for sharing of resources and, thus, simultaneous processing. Taken together, these models can describe how multitasking is handled differently by the brain based on preparation, complexity of tasks, and allocation of attention.

Preparation Neglect Hypothesis

Does this mean that simultaneous processing at the response selection stage is impossible? When multitasking appears in people’s daily lives, tasks tend to be much simpler and more familiar than those tested in PRP paradigm studies. For example, unlike novel laboratory tasks, talking to a friend while folding laundry is a well-practiced, familiar task that can often be performed with minimal demands on central processing resources. A study was conducted to address the concern of applying PRP paradigm studies to multitasking in daily life, and the results reveal an unexpected finding [13]. In one experiment, participants practiced tone (Task 1) and shape (Task 2) identification individually, with an emphasis on shape identification, and practiced multitasking before the experimental trials were conducted. The results of this experiment showed the typical bottleneck effect, with slower reaction times for Task 2 when the stimuli were presented close together [13]. Initial practice of Task 2 and multitasking scenarios may have improved overall reaction times, but the delay caused by stimulus overlap still remained. Similarly, an experiment tested whether practice could lead to a decrease in the Task 2 delay, which would mean less demand on executive function and possible bypassing of the bottleneck. The findings revealed that while slight improvements were observed, reaction times were still lower for tasks presented simultaneously [14]. This experiment further indicates that practice and familiarity alone are not always sufficient to efficiently multitask without task-switching.

On the other hand, when trials in which participants only completed Task 2 were intermixed with multitasking trials to increase participants' preparation or readiness for Task 2, 49.4% of the dual-task trials were consistent with bottleneck bypassing [13]. This means that when the stimulus onset asynchrony is very short, the delay seen in typical PRP paradigm studies was minimized by preparation for Task 2. This data supports the preparation neglect hypothesis, which proposes that dual-task interference, the increased reaction time for Task 2 during multitasking, occurs because of a lack of preparation for Task 2 [13]. According to this study, preparation enables simultaneous or parallel processing by adequately emphasizing Task 2. In other words, the brain is equally ready to complete Task 1 and Task 2, instead of considering tasks one at a time. These findings suggest that we may be able to multitask, but we are bad at preparing for the second task while doing the first.

Long-Term Effects of Multitasking

Whether evaluated with bottleneck models or capacity-sharing models, the short-term effects of multitasking are delayed decision-making and decreased reaction time. However, there are long-term effects of multitasking, particularly media multitasking. Media multitasking is the simultaneous use of two or more forms of media and has become a common practice among many younger individuals [10, 15, 16]. Some examples of this practice are texting a friend while watching a movie or show, listening to a podcast while scrolling on social media, and online shopping during a lecture. The standard measure of the frequency of media multitasking in one’s daily life is the Media Multitasking Index (MMI) [17]. An experiment was done to determine the ability of chronic multitaskers to ignore irrelevant information using the “2-Back” and “3-Back” tests [15]. Participants were asked to remember a target letter while other letters appeared on screen for multiple trials. Then, they were asked to determine whether the current letter was the one that appeared two or three letters ago. The results showed that those who engage in media multitasking more often in their everyday lives have more difficulty with this task as opposed to those who rarely engage in media multitasking. These findings point to the long-term impacts of media multitasking: a chronic state of distraction. Two main hypotheses try to explain this phenomenon. The “ability hypothesis” claims that individuals are less able to disregard irrelevant information after engaging in media multitasking frequently. This can be seen in a student who constantly gets distracted by notifications on their phone when trying to study. On the other hand, the “strategic hypothesis” brings to light the possibility that heavy media multitaskers cognitively choose to mentally attend to distracting information because they may need greater stimulation to engage with the task at hand. This can be seen in a student who must play a TV show in the background while studying. Both hypotheses suggest that chronic media multitasking may make it harder for people to complete specific tasks without distraction or to focus on relevant information [15].

Have you ever walked into a room and forgotten what brought you there in the first place? Whether it’s seeing something unusual while walking to the room or stopping to talk to a friend, distracting stimuli can completely throw us off from the task at hand if we regularly engage in heavy media multitasking. These seemingly simple behaviors are manifestations of brain changes caused by media multitasking. Studies have shown that when completing a task with a distracting stimulus, individuals who engage in higher levels of media multitasking have greater activity in the right prefrontal cortex. This region of the brain showed increased activity as task difficulty increased, indicating that the right prefrontal cortex is involved in attention and executive control [11]. This means that the right prefrontal cortex plays a major role in deciding what stimuli to pay attention to in order to complete one’s goals. Heavy media multitaskers had to exert more cognitive effort to complete a task with a distractor present compared to those with lower levels of media multitasking in their daily life [11]. It is important to recognize that although multitasking behaviors may seem trivial in the moment, they can have long-lasting negative physical effects. Likewise, small changes in our daily media multitasking habits can lead to significant long-term positive impacts.

Multitasking may seem appealing to many people, but examining the brain processes involved in attempting to accomplish tasks simultaneously reveals the costs of this seemingly effective practice.  So next time you find yourself saying, “I have too much to do and not enough time,” understand that multitasking has significant limitations, especially when it comes to complex tasks. Therefore, consider which tasks might deserve more than your shared attention. Our brains aren’t designed for the type of efficiency that we demand from them by attempting to complete more than one task at once. The bottleneck and central resource sharing theory exemplify how multitasking often forces the brain to rapidly switch between tasks. For students in particular, this can reframe the idea of productivity. Productivity is not about how many tasks can be juggled at one time, but about minimizing unnecessary mental load and demand on the brain. 

REFERENCES:

1. May, K. E., & Elder, A. D. (2018). Efficient, helpful, or distracting? A literature review of media multitasking in relation to academic performance. International Journal of Educational Technology in Higher Education, 15(1), 13. https://doi.org/10.1186/s41239-018-0096-z

2. Klapp, S. T., Maslovat, D., & Jagacinski, R. J. (2019). The bottleneck of the psychological refractory period effect involves timing of response initiation rather than response selection. Psychonomic Bulletin & Review, 26(1), 29–47. https://doi.org/10.3758/s13423-018-1498-6

3. Mittelstädt, V., Mackenzie, I., & Miller, J. (2022). Evidence of resource sharing in the psychological refractory period (PRP) paradigm, 48(11), 1279–1293. https://doi.org/10.1037/xhp0001052

4. Otermans, P. C. J., Parton, A., & Szameitat, A. J. (2022). The working memory costs of a central attentional bottleneck in multitasking. Psychological Research, 86(6), 1774–1791. https://doi.org/10.1007/s00426-021-01615-1

5. Wu, Y., & Wang, Q. (2024). Dual-task interference: Bottleneck constraint or capacity sharing? Evidence from automatic and controlled processes. Attention, Perception, & Psychophysics, 86(3), 815–827. https://doi.org/10.3758/s13414-024-02854-1

6. Karlin, L., & Kestenbaum, R. (1968). Effects of Number of Alternatives on the Psychological Refractory Period. Quarterly Journal of Experimental Psychology, 20(2), 167–178. https://doi.org/10.1080/14640746808400145

7. Pashler, H., & Christian, C. L. (1994). Bottlenecks in planning and producing vocal, manual and foot responses. 

8. Pashler, H., & Johnston, J. C. (1989). Chronometric Evidence for Central Postponement in Temporally Overlapping Tasks. The Quarterly Journal of Experimental Psychology Section A, 41(1), 19–45. https://doi.org/10.1080/14640748908402351

9. Wu, T., Spagna, A., Mackie, M.-A., & Fan, J. (2023). Resource sharing in cognitive control: Behavioral evidence and neural substrates. NeuroImage, 273, 120084. https://doi.org/10.1016/j.neuroimage.2023.120084

10. Carrier, L. M., Rosen, L. D., Cheever, N. A., & Lim, A. F. (2015). Causes, effects, and practicalities of everyday multitasking. Developmental Review, 35, 64–78. https://doi.org/10.1016/j.dr.2014.12.005

11. Moisala, M., Salmela, V., Hietajärvi, L., Salo, E., Carlson, S., Salonen, O., … Alho, K. (2016). Media multitasking is associated with distractibility and increased prefrontal activity in adolescents and young adults. NeuroImage, 134, 113–121. https://doi.org/10.1016/j.neuroimage.2016.04.011

12. Broeker, L., Ewolds, H., de Oliveira, R. F., Künzell, S., & Raab, M. (2021). The impact of predictability on dual-task performance and implications for resource-sharing accounts. Cognitive Research: Principles and Implications, 6(1), 1. https://doi.org/10.1186/s41235-020-00267-w

13. Lyphout-Spitz, M., Maquestiaux, F., Ruthruff, E., & Chaloyard, S. (2024). Uncorking the central bottleneck: Even novel tasks can be performed automatically. Journal of Experimental Psychology: Human Perception and Performance, 50(1), 74–98. https://doi.org/10.1037/xhp0001169

14. Strobach, T., & Schubert, T. (2017). No evidence for task automatization after dual-task training in younger and older adults. Psychology and Aging, 32(1), 28–41. https://doi.org/10.1037/pag0000144 

15. Ralph, B., & Smilek, D. (2016). Individual differences in media multitasking and performance on the n-back, 79, 582–592. https://doi.org/10.3758/s13414-016-1260-y

16. Uncapher, M. R., K. Thieu, M., & Wagner, A. D. (2016). Media multitasking and memory: Differences in working memory and long-term memory. Psychonomic Bulletin & Review, 23(2), 483–490. https://doi.org/10.3758/s13423-015-0907-3

17. Ophir, E., Nass, C., & Wagner, A. D. (2009). Cognitive control in media multitaskers. Proceedings of the National Academy of Sciences, 106(37), 15583–15587. https://doi.org/10.1073/pnas.0903620106