App Graveyard Entry #1: DoTo - A todo app
Positive Behavioral Reinforcement Mechanisms Embedded Within a Spatially Structured Digital Environment for Consumer Mobile Applications
what's in your pond?
how heavy does it feel?-
App Graveyard — Concept No. 1 Mar 17, 2026 · by Steve Brait
Abstract
DoTo was a consumer mobile application in which users created digital artifacts representing cognitive objects and tasks occupying their attention, where each artifact was given a size determined entirely by the user's own assessment of its weight relative to their cognitive capacity at that moment, and where those artifacts were placed as living entities inside a bounded digital environment designed to correspond visually to the cognitive state the user had declared through the act of creating them.The user's relationship with each artifact in the digital space was left open by design, in that an artifact could be removed when the user determined its associated concern was no longer relevant to them, or it could remain in the pond as a persistent presence. The digital environment did not prescribe what the user should do with what was in it, instead it gave the contents of the user's own ideation a visual form and a place to exist.The application was built on iOS using Swift, SwiftUI, and Core Data, completed to a functional prototype state, and never released to market. The reinforcement mechanisms embedded in it are documented here against their clinical foundations, and the application is evaluated against a consolidation framework that determines whether a discrete tool warrants its own infrastructure or belongs inside a larger system.
I. The Instrument
Building a new kind of todo application requires stepping outside the conventions of todo applications, because framing the design problem as an improvement to the list constrains the solution to the list's assumptions, its vocabulary, its neurological register, and its behavioral limitations. The design of DoTo began by abandoning that frame and asking a different question: What if the things on your mind were not added to a list but given form as living things , and placed into a space you could see?The cosmetic and functional implications of that reframe are inseparable, in that when a user thinks about how many items they have on a list, they are operating inside a cognitive structure that is discrete, enumerable, and administrative, whereas when a user thinks about how many fish they have in a pond, they are operating inside a different structure that is spatial, living, and environmental. The digital environment they inhabit, the language they use to describe what is on their mind, and the mental model they form of what it means to engage with those things all change together, and therefore my choice for the digital space and language I would embody was the metaphor of Fish in a Pond.Within that decision, I also chose to target a specific demographic: what I would call the Girly Girl segment. This was not a casual aesthetic direction, instead it was a deliberate scoping decision made with the idea in mind that this demographic offered the clearest and most clinically distinct design philosophy available, one with well-defined color psychology associations, a specific affective symbology palette, and a psychological profile that is systematically underserved by the aggressive, list-driven productivity tools that dominate the category. Every subsequent design decision in the fish skin, where skin refers to the complete visual and behavioral design theme applied over the underlying architecture, analogous to a palette or costume that changes the emotional register of the experience without altering the mechanisms beneath it, including the palette, the entry mechanism, the affirmation language, the size of the pond, and the visual register of the fish themselves, was constructed against that profile with the same clinical intention applied to each element individually.A fish in the pond is not necessarily a task to be completed, rather it is a digital artifact the user created to embody any cognitive object or task occupying their attention: a worry, a commitment, something unresolved, something important that has not yet been addressed, or simply something the user wants to hold in awareness. The size of that artifact was not inferred or assigned by the application, instead it was declared by the user at the moment of creation as a direct expression of the weight they attributed to that object relative to their cognitive capacity at that point in time, and whether an artifact stays in the pond, gets removed, or remains as a persistent presence is entirely a function of the user's own relationship with what it represents.The implementation was a population of fish inhabiting a circular digital pond on the device screen, where each artifact persisted in the digital environment for as long as the user chose to keep it there. The pond filled as the user's created artifacts accumulated, and its population at any given moment reflected the sum of what the user had declared to be occupying their attention. The fish swam, applying behavior to thought, with bigger fish naturally imposing their presence and drawing attention while smaller fish exhibited schooling behaviors, swimming together when their associated concerns were related, and the relative social dynamics of the pond, in which a large fish commands the space while smaller ones cluster and defer, was designed to mirror the way cognitive objects of different weights actually compete for attention in working memory. This concept drew direct inspiration from the ancient mnemonic technique known as the method of loci, or memory palace, in which information is mentally placed in specific physical locations within an imagined space, producing dramatically higher recall fidelity than abstract memorization alone (Yates, 1966), because the hippocampus, which manages both spatial navigation and memory consolidation, uses place as an organizing framework for information storage such that things that exist somewhere are tracked by the brain differently than things that exist only as text (O'Keefe and Nadel, 1978). The pond was my digital implementation of that principle: a located, living space in which cognitive objects were given bodies, positions, and behavioral presence, such that looking at the pond would feel qualitatively closer to the actual experience of carrying that load than looking at a numbered list does.When a user chose to remove an artifact, the corresponding entity left the digital space and triggered an immediate positive reinforcement signal calibrated to the user's demographic profile, and when the digital space was fully cleared, it triggered a sustained closure state representation designed to allow the neurological decompression of complete cognitive loop resolution to register fully rather than being immediately displaced by the next concern.II. The Architecture: Why the Digital Representation of Attributed Cognitive Load Matters
The dominant paradigm in task management is the list, and the list presents a structural problem that the category has largely not addressed, which is that it is a static digital representation of a dynamic cognitive state, and the two things do not correspond well.When a user is carrying things in their mental space, those things are not sitting quietly in a column, instead they are running as background processes in working memory, surfacing at inconvenient moments, competing for attentional resources, and producing a diffuse cognitive pressure that persists independently of whether the user is actively thinking about any of them.A numbered list on a screen does not communicate that, because it presents lines of text that the brain processes as lines of text rather than as a representation of active cognitive processes running in parallel, and the gap between the digital representation and the cognitive reality it is meant to reflect can itself become a source of low-level stress, given that the tool presents something that does not correspond to the user's actual experience of carrying the load.The architecture of DoTo was built to close that gap by constructing a digital environment whose visual state at any given moment would correspond more closely to the cognitive state the user had declared through the act of creating their artifacts, where the user's ideations were given form as living entities placed inside a bounded digital space, sized according to the user's own attribution of their weight relative to their cognitive capacity. The space filled as those artifacts accumulated, and when the user chose to remove an artifact, the digital environment changed in direct response, such that an empty digital space corresponded to the user having declared nothing left to carry.That empty state was not treated as an absence, instead it was designed as a destination, in that when the final artifact left the pond, the pond itself did not simply disappear but instead breathed, pulsing gently before fading over thirty seconds, then holding a beat of empty screen before the heart returned. The psychology of that sequence is grounded in the same closure research that underlies the removal mechanism itself, in that Zeigarnik's work implies the cognitive benefit of resolution is most fully realized when the closed loop is given time to complete before the next one opens (Zeigarnik, 1927), and a tool that immediately replaces the empty state with new content denies the brain the decompression it has earned. The fade was designed to let the user feel the weight they had been carrying by briefly letting them feel its absence, which is a distinct and clinically meaningful neurological event, and one that the vast majority of productivity applications have never attempted to represent. This sets the stage for understanding why the entry state of the application, the moment before any cognitive load has been declared at all, required the same level of intentional design.Every variable in that structure is replaceable except the core relationship between user-created artifact, digital space, and user agency over what occupies it, where the fish skin is one implementation of that relationship and other skins are structurally available within the same architecture. Robots dispatched from a gridded digital environment give a futuristic, efficient register, zombies advancing across a yard build urgency into the visual metaphor because clearing them is defense rather than administration, monsters in a dungeon give a fantasy vocabulary where the artifacts in the space become creatures the user defeats, and asteroids in orbit give an isolating scale where the weight of what the user has placed in the digital space is built into the image itself.Each of these is a different skin on the same architecture, and each requires its own demographic and neurological calibration before the reinforcement mechanisms it carries can function as designed, because the choice of skin determines which neurological and psychological systems are recruited to deliver reinforcement, and recruiting the wrong systems for the wrong user does not merely reduce effectiveness but can actively work against the behavioral goals the architecture was built to support.III. Mechanism No. 1: Affective State Priming at Application Entry
The entry state of most consumer software is treated as functionally inert, as though nothing clinically significant happens in the moment before the product's main function is engaged, but research on affective priming suggests otherwise.The emotional tone of the first stimulus a brain encounters in a given context shapes how it processes everything that follows (Fazio, 2001), and the digital entry state of an application therefore establishes an emotional register that either supports or works against the user's relationship with the tool, which in productivity software specifically has measurable consequences for whether the tool gets used at all.The primary failure mode of most task managers is that users stop opening them, because repeated associations between opening the tool and the cognitive experience of confronting obligation, pressure, and the accumulated weight of everything on the mind produce an aversive association that, once established through repeated pairing, causes the behavior of opening the tool to undergo extinction through classical conditioning before any of its mechanisms have a chance to function (Norman, 2004).Most productivity applications display an empty list at the zero state, which establishes a neutral-to-negative emotional register before anything has been added, whereas DoTo displayed a heart with a plus sign. The design of that entry element was both cosmetic and functional, in that I constructed it as a dual-channel stimulus with the idea in mind of engaging two distinct neurological systems simultaneously.The plus sign operates as an objective additive symbol engaging the prefrontal cortex through its association with construction, addition, and forward action, communicating that something is being created and that creation is within the user's agency, while the heart operates as a subjective affective symbol engaging the hypothalamus through its established association with warmth, safety, and positive valence. For the target demographic, a user who carries a preloaded positive association with the aesthetic register of pink hearts, the hypothalamic response to that symbol is not neutral, instead it is a conditioned signal that this space is a good one to be in.The combination of the two symbols in a single entry element means the moment of opening the application and initiating the creation of a new artifact delivers both a cortical readiness signal and an affective safety signal at the same time, such that the user is primed to act and primed to feel safe acting before a single artifact has been created. That is an affective priming mechanism at application entry, and its function is to interrupt the formation of the aversive association before it consolidates.IV. Mechanism No. 2: User-Attributed Cognitive Weight Encoding at Artifact Creation
Most task managers allow users to add entries without consciously processing what those entries represent, where the user types the entry, moves on, and the cognitive weight of what was just logged goes unassessed, and the digital representation treats it identically to every other entry regardless of what it actually costs, with the result that the visual state of the list communicates nothing about the actual cognitive burden the user is carrying.Ten entries could represent ten trivial concerns or ten things causing significant stress, and the list looks identical either way, which is the core structural failure that research on cognitive load theory speaks to directly, in that externalizing and organizing cognitive load into structured representations reduces the attentional cost of managing that load over time (Sweller, 1988), and a representation that assigns identical visual weight to all entries regardless of their actual cost to the user does not externalize load in any meaningful sense.DoTo introduced a size selection step that the user completed before any artifact entered the digital pond: small, medium, or large, where the attribution belonged entirely to the user and no algorithm inferred it and no system imposed it. The user looked at what they were about to give form to, made a judgment about its weight relative to their cognitive capacity at that moment, and expressed that judgment through their size selection, and that act of self-attributed sizing is what researchers call a metacognitive assessment, a conscious evaluation of cognitive weight, where placing that attribution on the user rather than the system is a deliberate design choice made with the understanding that only the user knows what a given cognitive object actually costs them on a given day.A pond containing three large fish and one small fish looks and feels different from a pond containing four medium fish, and that difference is a direct reflection of the user's own declared assessment of their cognitive load, where the size encoding makes that attribution visible and persistent in the digital space in a way the list does not, and the act of attributing size requires the user to engage consciously with what they are creating before it enters the pond.V. Mechanism No. 3: Visuospatial Encoding of User-Attributed Load Through Hippocampal Organization
The memory palace works because of hippocampal organization rather than because of the technique itself, in that the hippocampus manages both spatial navigation and memory consolidation by using physical location as an organizing framework for information storage, such that items associated with a specific place are more robustly encoded and more readily retrieved than items held in purely abstract form (O'Keefe and Nadel, 1978). The method of loci, the ancient technique of mentally placing information in imagined physical locations, produces dramatically higher recall fidelity than abstract memorization for exactly this reason (Yates, 1966).Baddeley's model of working memory includes a visuospatial sketchpad dedicated to processing and temporarily holding visual and spatial information (Baddeley, 2000), and a text-based list engages this component minimally because the words are processed as symbols rather than as objects in a located digital space.A circular digital pond populated with the user's own created artifacts at specific positions, each carrying the size and color the user attributed to them and each exhibiting the swimming behaviors described in Section I, engages the visuospatial sketchpad directly and persistently, because the brain processes each artifact as an object in a located digital space rather than as a symbol in a list. The result is that the artifacts track as present things in the user's awareness more persistently than list entries do, given that the hippocampal and visuospatial systems monitoring their status in the digital space are older, more automatic, and less dependent on executive effort than the systems required to consciously monitor a text list.VI. Mechanism No. 4: Analog Digital Representation of Zeigarnik Cognitive Load
In 1927, Soviet psychologist Bluma Zeigarnik documented that people recall interrupted or incomplete things significantly more accurately than completed ones (Zeigarnik, 1927), with the interpretation being that the brain maintains an active cognitive loop for each unresolved item in awareness and keeps that loop running until it receives a resolution signal. This is the Zeigarnik effect, and it explains why unresolved concerns surface in awareness during unrelated activities, given that the brain maintains awareness of what is unresolved until it is no longer unresolved.The standard text list does not represent Zeigarnik load accurately, because a list of ten items is ten words on a screen, whereas the cognitive experience of carrying ten simultaneously active Zeigarnik loops in working memory does not correspond to ten words on a screen, and the gap between what the digital tool shows and what the cognitive system is managing means the tool fails to reflect the actual weight of the load the user has declared.DoTo's digital pond represents that declared load rather than counting it, in that the digital space fills as the user creates more artifacts, a full pond corresponds visually to a state the user has declared to be full, and a nearly empty pond corresponds to a state the user has declared to be nearly clear. The swimming behaviors of the fish compound this effect in a meaningful way, because the user is not only prompted by the presence of the fish as cognitive objects in a located space but also by the metacognitive awareness of their activity, in that the fish are still moving, still swimming through the mental pond, and that continued motion signals to the observing mind that these concerns have not resolved themselves, they are not sitting still, they are ongoing and alive in exactly the way the Zeigarnik effect describes unfinished cognitive loops as being. The experience of watching the digital space and its living inhabitants change in direct correspondence with the user's own attributed cognitive state is therefore operating on both the spatial memory system and the biological presence-monitoring system simultaneously.VII. Mechanism No. 5: Digital Spatial Agency as a Strong Cognitive Closure Signal
Marking an item on a list leaves it in the digital space, where it acquires a new visual state, perhaps a strike-through or a muted color, but it continues to occupy the same digital location it occupied before, and the cognitive loop it was sustaining does not fully close because the object generating it is still present. The brain processes the annotation but the artifact remains, and the loop stays partially open.Research on the need for cognitive closure, the motivated desire for a definitive signal that an active cognitive process has genuinely ended rather than merely been marked, demonstrates that weak closure signals produce incomplete closure states that continue to consume cognitive resources even after the relevant behavior has been completed (Kruglanski and Webster, 1996), and a checkmark is a weak digital closure signal precisely because it updates the status of the artifact while leaving the artifact in place.When a user chooses to remove an artifact from the digital pond, the object sustaining the Zeigarnik loop no longer exists in the digital environment, the visuospatial sketchpad stops tracking it, and the cognitive loop closes because there is nothing left in the digital space to keep it open, such that the act of removal becomes the user's own declaration that the cognitive object this artifact represented no longer needs to occupy their space, and the digital environment responds to that declaration by making it literally true.The natural vocabulary English speakers already use for this kind of cognitive clearing is consistently spatial, in phrases like get it off your mind, clear your head, put it to rest, and the mechanism in DoTo implements that literally rather than leaving it figurative.VIII. Mechanism No. 6: Positive Reinforcement Signal Calibrated to Removal Events Rather than Attributed Significance
A common intuition in productivity design is that rewards should scale with the significance of what was accomplished, but applied as a calibration principle that produces a reinforcement schedule which provides no signal for the smaller or lower-weight artifacts that are most commonly accumulating in the digital space and sustaining the background cognitive load the architecture was designed to reduce.The habit of clearing things from one's mental space, practiced on small things through repeated reinforced execution, operates on the same behavioral substrate as the habit of clearing large things, and a reinforcement schedule that only fires for artifacts the user attributed high weight to fails to train the behavior at the level where it most needs training.Operant conditioning research demonstrates that a reinforcement signal strengthens the behavior it follows when delivered immediately and consistently after the target behavior occurs (Skinner, 1938), and the behavioral target in DoTo is the act of removing a user-created artifact from the digital space independent of the weight the user attributed to it, so that delivering a consistent positive reinforcement signal for every removal event regardless of artifact size trains the association between clearing the space and reward at the level of the behavior itself.The affirmation language that carries the reinforcement signal, warm and second-person phrases like "Crushing It!" rather than neutral confirmation messages, is a delivery mechanism calibrated to a specific neurological and demographic profile, and the reasoning behind that calibration is addressed in the following section because it is the component that changes most significantly across different skin implementations.IX. Mechanism No. 7: Demographic-Calibrated Affirmation Delivery and Neurological Skin Selection
Motivation is not a uniform quantity that users either possess or do not possess, instead it is a system of neurological processes that vary significantly across individuals and respond to different stimuli in measurably different ways. The behavioral outcomes of a positive reinforcement mechanism depend on whether the signal it delivers matches the neurological systems of the user receiving it, and the correct match is determined by the user's demographic profile, psychological baseline, and the cortical pathways the skin has been designed to engage.The fish skin engages reinforcement systems in users who respond to warmth, verbal encouragement, and low-threat environmental framing, where self-affirmation theory, developed by Steele (1988), holds that affirming a core aspect of the self-concept functions as a buffer against threat-induced defensiveness and restores the individual's sense of self-integrity. When DoTo delivers "Crushing It!" at the moment a user removes an artifact from the pond, it affirms the user's identity as someone who manages the contents of their mental space rather than commenting on the specific cognitive object that was removed.For users whose baseline self-efficacy around managing their cognitive load is lower, where self-efficacy is the belief in one's capacity to execute the behaviors required to produce a specific outcome (Bandura, 1977), this identity affirmation at the moment of removal carries measurable behavioral weight by strengthening the association between the clearing behavior and a positive self-image.Children are a particularly responsive demographic for this mechanism, given that the dopaminergic circuits mediating response to social praise and immediate verbal positive reinforcement are highly sensitive during childhood and modulate behavior more immediately than abstract future-outcome reasoning, which requires prefrontal development not complete until early adulthood (Casey et al., 2008). An affirmation delivered at the moment a user removes an artifact from the pond is near-optimal reinforcement delivery for a child user because the signal is immediate, emotionally salient, and socially coded in a way the developing reward system responds to strongly.Users with higher trait anxiety, higher neuroticism, or chronic low self-efficacy show the same preference for warm and affirming language in completion contexts, with one important constraint identified by Wood et al. (2009) being that positive self-statements can backfire for users with low self-esteem when those statements feel incongruent with the user's actual self-assessment. The affirmation language in DoTo is therefore calibrated to the specific act rather than the global self, where "You crushed it" attaches to a behavior the user just demonstrably performed in the digital space, whereas "You are amazing" makes a claim about the person that a user with a low self-concept may reject, converting the reinforcement mechanism into an aversive one.The zombie skin engages a different set of systems entirely, in that a user clearing zombie artifacts from their digital yard is responding to a threat, removing it, and receiving the neurological relief that follows threat resolution, where the anterior cingulate cortex and amygdala are more centrally involved through their roles in threat-monitoring and threat-resolution, and the dopamine response is organized around threat removal rather than self-concept reinforcement. A competitive, high-drive user who experiences affirmation language as condescending would respond well to zombie urgency because the digital environment frames clearing the space as territorial defense rather than administrative self-care, and the same spatial architecture running a different positive reinforcement mechanism through a different cortical pathway for a different user is why the choice of skin is a clinical design decision rather than a visual preference.X. Mechanism No. 8: Sustained Full Closure State Representation
A partially cleared digital space and a fully cleared digital space are not the same cognitive event and do not produce the same neurological state, yet most productivity applications treat them identically by returning the user to an empty list either way, which means the state of having no remaining artifacts in the space, the simultaneous absence of every active Zeigarnik loop associated with the digital environment, goes unrepresented.Zeigarnik's original research implies that the cognitive benefit of clearing is most fully realized when the closed loop is given time to complete its resolution before the next loop opens (Zeigarnik, 1927), and when a digital application immediately replaces an empty environment with new content, that state barely registers before reopening, and the distinctive positive reinforcement it produces is never fully delivered.When DoTo's final artifact was removed from the digital pond, the pond pulsed gently, then faded over thirty seconds, then held an empty digital screen for a beat, and then the heart returned. That sequence was designed to hold the full empty state of the digital environment open long enough for the brain to register that nothing remains, to receive the signal that the space is clear, and to associate that signal with the behavior that produced it. That is closure, and designing for it explicitly rather than treating it as an absence was the point.XI. Mechanism No. 9: Palette Selection as Aversive Conditioning Countermeasure
The documented failure mode of most productivity tools is that users stop opening them, because repeated pairing of the tool's visual identity with the cognitive experience of confronting obligation, pressure, and the accumulated weight of everything on the mind causes the digital tool itself to become a conditioned aversive stimulus, and the behavior of opening it undergoes extinction (Norman, 2004).A digital color palette that reads as gentle, warm, and non-threatening disrupts the formation of that aversive association before it consolidates, and the fish specifically, soft and non-aggressive and present in their digital pond, give the user's own created artifacts a visual form that does not trigger the low-level threat response that lists of unfinished things typically generate.This is a precondition mechanism rather than a reinforcement mechanism, and its function is to prevent the digital application from becoming something the user's brain learns to avoid, so that every positive reinforcement mechanism in the architecture has the opportunity to operate.XII. Project Consolidation
Any discrete application that can deliver its full intended function as a component within a larger system does not warrant existence as a standalone product, and this is a principle I apply to design, which I call Project Consolidation.A standalone application carries infrastructure overhead, organizational reach requirements, maintenance cost, and what I call E-Space, which is the existential cognitive and attentional territory a product must occupy in the world to justify its own presence, and if the positive reinforcement mechanisms embedded in a digital application can operate inside a larger container without losing their effectiveness, then building them outside that container spends all of those resources without corresponding gain.Project Consolidation is not a judgment about whether an idea is good, instead it is a judgment about the correct form for a good idea to take.DoTo does not pass that evaluation, because the mechanisms it contains, visuospatial encoding of user-attributed cognitive load in a digital space, demographic-calibrated affirmation delivery, and user agency over the digital environment as a strong cognitive closure signal, do not require their own App Store presence, or their own E-Space, outside of this post.XIII. Summary of Findings
Every positive reinforcement mechanism embedded in DoTo was pointed at a specific and documentable psychological reality, grounded in research spanning nearly a hundred years from Zeigarnik's 1927 observations about incomplete cognitive loops through Baddeley's working memory model through Fogg's behavioral design framework.The fish skin was the right first implementation because the demographic that responds most strongly to identity-affirming language and low-threat digital environmental design is also the demographic least well-served by existing productivity tools, where children, high-anxiety users, users with chronic overwhelm, and users who need the act of clearing their space to carry its own immediate positive signal are not well-served by red priority flags and streak counters. The prototype established that the mechanisms work for those users and can be recalibrated for others without altering the digital architecture that houses them.Doto will now live forever Here, as a shadow of itself.
Rest in pixels.— Steve
References
Baddeley, A. D. (2000). The episodic buffer: A new component of working memory? Trends in Cognitive Sciences, 4(11), 417–423.Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84(2), 191–215.Casey, B. J., Getz, S., and Galvan, A. (2008). The adolescent brain. Developmental Review, 28(1), 62–77.Fazio, R. H. (2001). On the automatic activation of associated evaluations: An overview. Cognition and Emotion, 15(2), 115–141.Fogg, B. J. (2009). A behavior model for persuasive design. Proceedings of the 4th International Conference on Persuasive Technology (Article 40). ACM.Kruglanski, A. W., and Webster, D. M. (1996). Motivated closing of the mind: "Seizing" and "freezing." Psychological Review, 103(2), 263–283.Norman, D. A. (2004). Emotional design: Why we love (or hate) everyday things. Basic Books.O'Keefe, J., and Nadel, L. (1978). The hippocampus as a cognitive map. Oxford University Press.Skinner, B. F. (1938). The behavior of organisms: An experimental analysis. Appleton-Century-Crofts.Steele, C. M. (1988). The psychology of self-affirmation: Sustaining the integrity of the self. Advances in Experimental Social Psychology, 21, 261–302.Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257–285.Wood, J. V., Perunovic, W. Q. E., and Lee, J. W. (2009). Positive self-statements: Power for some, peril for others. Psychological Science, 20(7), 860–866.Yates, F. A. (1966). The art of memory. University of Chicago Press.Zeigarnik, B. (1927). Über das Behalten von erledigten und unerledigten Handlungen [On the retention of completed and uncompleted actions]. Psychologische Forschung, 9, 1–85.