Heterocrasis

1A different starting mix

The earlier papers in this series describe how media and drugs change the configuration of the senses. Sensory Syncrasis proposed that configuration has five dimensions: the relative weight of each sense, the overall gain applied to all of them, whether signals are fused or kept apart, the balance between signals from the body and from the world, and how long a setting persists. Typomora and Pharmakrasis traced how keyboards, voices and drugs shift those settings.

All of that assumes a starting point, and starting points differ. Some differences are small and individual, the ordinary idiosyncrasy the series takes its name from. Others may be large and systematic enough to shape how a person meets every medium. This paper asks whether ADHD is one of them.

We call a systematically different starting configuration a heterocrasis. The question is whether ADHD is a heterocrasis, and if so, on which dimensions.

The answer here is partial. One dimension, overall gain, has good support from several independent lines of evidence. One, bodily self-perception, has mixed support undermined by how it has been measured. One, the integration of the senses, points the opposite way from what a simple story would predict. And one older idea, that ADHD involves deficient inner speech, sits poorly with the newer evidence and needs to be reframed.

Two framing choices matter. First, this is not a deficit account. A heterocrasis is a different setting, not a broken one, and some of what looks like impairment in one environment may fit another better. Second, ADHD is a diagnosis that covers a great deal of variation. Claims here describe group tendencies in research samples, not any individual. The aim is to locate ADHD on the map the series has drawn, and to be honest about where that map is still blank.

2Gain: the best-supported difference

Overall gain is how strongly the brain amplifies incoming signals relative to background noise. It is set largely by catecholamines, the neuromodulators noradrenaline and dopamine, and in particular by the locus coeruleus, a small brainstem nucleus that broadcasts noradrenaline across the cortex (Aston-Jones & Cohen, 2005)1. This is the dimension on which ADHD differs most clearly.

The theory

The idea that ADHD involves under-arousal is old. Zentall's optimal stimulation theory proposed that people with ADHD seek stimulation because their baseline arousal is too low for comfort. Sikström and Söderlund's moderate brain arousal model (2007)3 made it more precise. It argued that low dopamine levels in ADHD leave the nervous system short of the background noise needed for weak signals to register. Adding external noise could then help, through a phenomenon called stochastic resonance, in which a moderate amount of noise lifts a faint signal over the threshold of detection.

A computational account by Hauser, Fiore, Moutoussis and Dolan (2016)4 framed ADHD in terms of impaired neural gain across levels of analysis: weaker amplification of relevant signals relative to noise, traced to catecholamine systems.

The evidence

The noise prediction has been tested. A 2024 meta-analysis of 13 studies with 335 participants found that white or pink noise gave a small benefit on lab attention tasks to young people with ADHD or high attention problems (g = 0.25), and a small cost to young people without (g = −0.21) (Nigg and colleagues, 2024)5. The opposite direction of the effect in the two groups is what a gain account predicts. The benefit is small, the authors note concerns about control conditions, and one well-controlled study found no effect of noise on eye-movement control in children with ADHD (Jostrup and colleagues, 2024)6.

Computational studies of decision-making point the same way. In a reversal-learning task, adolescents with ADHD learned at the same rate as controls but chose more randomly, with a higher "decision temperature" and weaker prediction-error signals in the medial prefrontal cortex (Hauser and colleagues, 2014)7. A review of computational models of ADHD found that slower evidence accumulation is the most consistent finding, and that lower sensitivity in choice is better supported than altered learning rates (Ziegler and colleagues, 2016)8. Noisier choice with intact learning is what weak gain would produce.

Stimulant medication, which raises catecholamine levels, is among the most effective treatments for ADHD symptoms in large comparative analyses (Cortese and colleagues, 2018). Work on prefrontal pharmacology, associated especially with Arnsten, describes stimulants as improving the signal-to-noise ratio of prefrontal neurons. That fits the gain picture, though it should be stated with care: a drug helping does not prove what the condition is.

What it means

In the terms of this series, the clearest feature of the ADHD configuration is a gain setting that leaves weak signals struggling against noise. Much of what follows, including the hunger for stimulation and the benefit of background noise, can be read from there.

3The body: mixed evidence

Interoception, the perception of the body's internal signals, is the dimension where ADHD research has grown fastest and where the evidence is least settled.

What the studies found

The first systematic review of interoception in ADHD covered 17 studies. Of the five that compared people with ADHD against controls, three found reduced interoception; population studies linked lower interoception to inattention, impulsivity and emotional dysregulation (Bruton and colleagues, 2025)10.

The individual studies disagree. In one, 14 adults with ADHD were worse than 16 controls at counting their own heartbeats (Kutscheidt and colleagues, 2019)11. In another, 24 adults with ADHD and 23 controls performed "strikingly similar\[ly\]" (Wiersema & Godefroid, 2018)12. The largest case-control study, comparing 30 adults with ADHD against 57 controls, found lower accuracy, lower confidence and lower self-reported interoceptive ability in ADHD, even after accounting for autistic traits, alexithymia, depression and other factors (Göz Tebrizcik and colleagues, 2025)13.

The measurement problem

All three adult case-control studies used heartbeat counting, in which people silently count their heartbeats over a set period. That task is widely criticised, because scores are strongly influenced by what people believe about their own heart rate and by their ability to estimate time (Zamariola and colleagues, 2018)14. Since time estimation itself can differ in ADHD, the confound is especially serious here.

A large 2026 study in Nature Mental Health, using better-designed tasks in hundreds of participants, found objective interoceptive performance largely unrelated to mental-health symptoms, including ADHD traits. And a study of adolescents found that those with more ADHD symptoms showed stronger, not weaker, brain responses to their own heartbeats (heartbeat-evoked potentials).

What it means

The honest reading is that bodily self-perception may differ in ADHD, but the direction and size of the difference are unclear, and the measurement method behind most positive findings is the weakest available. The adolescent brain data suggest the body's signals may be processed differently rather than simply registered less. For this paper, the inside–outside balance is a hypothesis about ADHD, not a finding. A study using heartbeat discrimination rather than counting, with heart rate and time estimation controlled, is the obvious gap in the literature.

4Binding: integration is not looser

A tempting story about ADHD says the senses are poorly bound together: sights and sounds drift apart, the world arrives in fragments, and attention scatters as a result. The evidence does not support it.

The brain fuses a sight and a sound into one event if they arrive within a certain window of time, the "temporal binding window." A looser integration would mean a wider window. Panagiotidi, Overton and Stafford (2017)17 tested adults with high and low ADHD-like traits and found the opposite: a significantly narrower window in the high-trait group, with no shift in the point at which sights and sounds seemed simultaneous. A clinical study found audiovisual integration in adults with ADHD to be robust, varying with perceptual load (Schulze and colleagues, 2022)18. A 2025 study found that impulsivity did not affect how people weighed audiovisual evidence in decisions.

What is well documented is that people with ADHD report more unusual sensory experiences. Questionnaire studies using the Sensory Profile find higher rates of both over-responsiveness and under-responsiveness to sensation in ADHD, in children and in adults (Ghanizadeh, 201120; Bijlenga and colleagues, 2017). These are self- and parent-reports, and they describe how sensation feels rather than how it is combined.

What it means

Integration mode in ADHD does not look loose. If anything, the one direct study suggests tighter temporal binding, and clinical integration appears intact. The reported sensory atypicality is real but may come from somewhere else, plausibly from gain: a nervous system whose signal-to-noise setting is unusual would experience sensation as too much in some moments and too little in others. That reading is this paper's interpretation. It suggests the sensory reports of ADHD belong under gain rather than integration, which is a more specific claim than the fragmented-world story and fits the evidence better.

5Inner speech: coupling, not quantity

The deficit idea

The most influential theory of ADHD, Barkley's (1997)21, placed a deficit in behavioural inhibition at the centre and argued that it delays the internalisation of speech. Following Vygotsky, inner speech develops from speech: children first talk aloud to guide themselves (private speech), then the talk goes silent. If that internalisation is delayed, the argument ran, the self-guiding voice is weaker, and so is self-regulation.

There was early support. Berk and Potts (1991)22 found that boys with ADHD used less mature, less internalised private speech than other boys. The idea that ADHD is partly a "regulatory vacancy," a missing inner voice, follows naturally from this.

What the newer evidence shows

The newer evidence complicates the picture. Children with ADHD produce more private speech, much of it unrelated to the task (Corkum and colleagues, 2008)23. Preschoolers at risk for ADHD use more private speech, less tightly linked to how well they perform (Winsler and colleagues). In two large non-clinical samples of adults, higher ADHD-like traits were associated with more self-directed and inner speech, not less (Benfield and colleagues, 2025)25.

Inner speech in adults with diagnosed ADHD has barely been studied. When Alderson-Day and Fernyhough reviewed the field in 201526, they wrote that, to their knowledge, there had been no studies specifically of inner speech in ADHD.

Emotion is related. Emotional dysregulation is now recognised as a core feature of ADHD (Shaw and colleagues, 2014)27. Alexithymia, difficulty identifying and describing one's own feelings, appears common: 22% in one German clinical sample with no control group, and 42% in a Turkish sample compared against controls. Naming feelings in words reduces the brain's reactivity to them in general populations (Lieberman and colleagues, 2007)29, but whether that works differently in ADHD has not been tested.

Reframing: a coupling, not a vacancy

The evidence does not describe a missing inner voice. It describes a voice that is present, sometimes louder than usual, but less tightly coupled to action. The self-talk happens; it does less of the steering. That is a different and more specific claim than the deficit account, and a better fit with the data.

In the terms of this series, inner speech is a medium the mind uses to regulate itself. Heterocrasis on this dimension is not an absent medium but a loosely coupled one. This connects to the argument of Typomora, which proposed that speaking ideas aloud to an attentive listener, including an AI, reverses Vygotsky's path and returns thinking to its social, spoken form. If inner speech in ADHD is present but weakly coupled, an external listener that asks for the full version and holds the thread might restore some of the coupling from outside. That is speculation, and nothing has tested it, but it follows directly from the evidence as reframed here.

6Heterocrasis across the five dimensions

The table brings the evidence together. It is interpretation: the five dimensions are this series' framework, and no study has measured all of them in ADHD.

Two further features sit beside the table because they concern regulation rather than the senses directly. Inner speech is present and often abundant but weakly coupled to action (section 5). Emotional regulation is a core difficulty, with alexithymia common. Both can be read as the regulatory side of the same configuration.

The resulting picture of heterocrasis is narrower than the strong version of the idea and, for that reason, more useful. ADHD is not a general disturbance of the senses. It is a specific setting of gain, with consequences for how sensation is experienced and how inner speech steers action.

7Media as external configuration

If ADHD involves a gain setting that leaves weak signals struggling against noise, an obvious question follows: do people with that setting use media to supply what their nervous system does not?

What is known

The best longitudinal evidence runs the other way, from media to symptoms. Among adolescents with no significant ADHD symptoms at the start, each additional high-frequency digital media activity was associated with slightly higher odds of developing symptoms over two years (Ra and colleagues, 2018)30. Among students who used seven of the activities often, 9.5% developed symptoms, compared with 4.6% of those who used none. The authors called the association modest and noted that causation was not established. Broader reviews find small associations between media use and ADHD-related behaviour, likely running in both directions (Beyens, Valkenburg & Piotrowski, 2018)31.

No study has directly tested whether people with ADHD use media as a source of stimulation to regulate their own arousal.

A reading from the gain account

The noise studies offer a precedent. If adding background noise helps a low-gain nervous system detect weak signals (section 2), then other external stimulation might act similarly: music while working, a video playing in the background, a fast-moving game, constant small inputs from a phone. On this reading, some heavy media use in ADHD would be configuration-seeking: using the environment to set the gain the body does not.

The same reading predicts limits. Stochastic resonance has an optimum. Moderate noise helps; too much drowns the signal. Media designed to capture attention, with notifications, autoplay and endless feeds, are not tuned to anyone's optimum. They may supply stimulation and overshoot it in the same moment.

Two media, two directions

The earlier papers suggest two different relationships. Media that provide stimulation without structure, such as feeds and notifications, may add gain but also add noise. Media that provide structure for thought, such as speaking ideas aloud to a listener that responds, may strengthen the weakly coupled inner speech described in section 5. The first works on gain from outside. The second may work on coupling from outside.

This section is this paper's own synthesis. It describes a possibility the gain evidence makes plausible, not a finding. But it suggests that, for a heterocratic nervous system, the media question is less about how much screen time and more about what kind of configuration a given medium supplies.

8Objections and replies

"ADHD is too varied to have one configuration." True. ADHD covers inattentive, hyperactive and combined presentations, and individuals differ widely. The claims here describe group tendencies in research samples. Heterocrasis names a direction many people with ADHD share, not a profile every one of them has.

"Gain is just a new name for under-arousal theory." In part, and the paper credits that lineage (Zentall; Sikström and Söderlund). What the syncrasis framework adds is placement: gain is one of five dimensions, and saying which dimension differs, and which do not, is more specific than saying arousal is low.

"ADHD already has better-established models: executive dysfunction, delay aversion." It does, and this paper does not replace them. Executive-function and delay-aversion accounts, including Sonuga-Barke's dual-pathway model, describe how people with ADHD plan, inhibit and respond to waiting. Heterocrasis concerns a different level: the configuration of the senses those processes run on. A weak gain setting is compatible with, and may contribute to, both executive and motivational difficulties. The claim is narrower: of the five sensory dimensions, gain is where ADHD differs most clearly.

"The noise effect is tiny." It is small (g = 0.25) and not universal. Its value for this paper is less its size than its direction: the same noise that helps people with ADHD slightly impairs people without it. That reversal is what a difference in gain predicts.

"The interoception findings are stronger than you allow." Three of five case-control studies did find reduced interoception, and the largest controlled for many confounds. But all used heartbeat counting, a task known to be confounded by beliefs about heart rate and by time estimation, and a large study with better tasks found little relationship. Treating the dimension as mixed is the fair reading until discrimination-based studies exist.

"Reframing inner speech as coupling rescues Barkley's theory with new words." It changes the theory more than it rescues it. Barkley predicted less internalised speech; the newer evidence finds as much or more. The coupling reading keeps the part of the theory that holds, that inner speech steers behaviour less effectively, and drops the part that does not.

"Calling it a different configuration rather than a disorder minimises real difficulty." The difficulties are real, and the paper does not deny them. A configuration that serves poorly in quiet classrooms and long reading tasks still produces genuine impairment there. Calling it a heterocrasis says where the difference lies and allows that some environments fit it better than others. It is not a claim that support or treatment is unnecessary.

"The media section is speculation." It is, and it is labelled as such. It is included because it follows directly from the gain evidence and because media are the subject of this series.

9Conclusion

The series this paper belongs to argues that media and drugs re-temper the configuration of the senses. That argument needs a starting point, and starting points differ. ADHD is the clearest case where the difference may be systematic.

The evidence supports a narrower heterocrasis than the strong version of the idea. ADHD does not look like a general disturbance of the senses. Integration is not looser; bodily self-perception is unclear; inner speech is present rather than missing. What stands out is gain: a nervous system in which weak signals struggle against noise, so that stimulation helps, choices run noisier, and sensation is experienced as too much or too little. Around that sits a regulatory feature: inner speech that is abundant but loosely coupled to action.

Read this way, much that is described as deficit becomes a mismatch between a configuration and an environment. A setting that struggles in a silent room with a long text may do well in a stimulating one, and the right external input, whether background sound or a listener who holds the thread, may supply what the configuration lacks.

The open questions are specific. Does interoception differ in ADHD when measured with discrimination tasks rather than counting? Does the benefit of noise follow the optimum that stochastic resonance implies, and does it depend on a person's baseline arousal? Is inner speech in adults with ADHD different in how it steers action, rather than in how much of it there is? And do people with ADHD use media to regulate their own gain, and does it help or overshoot? Each of these is a question about the configuration a person starts from, and how the media around them meet it.

References

1. 1Aston-Jones & Cohen (2005), Annual Review of Neuroscience 28: 403–450
2. 2Zentall & Zentall (1983), Psychological Bulletin 94: 446–471
3. 3Sikström & Söderlund (2007), Psychological Review 114(4): 1047–1075
4. 4Hauser, Fiore, Moutoussis & Dolan (2016), Trends in Neurosciences 39: 63–73
5. 5Nigg et al. (2024), JAACAP 63(8): 778–788
6. 6Jostrup et al. (2024), Journal of Attention Disorders
7. 7Hauser et al. (2014), JAMA Psychiatry 71(10): 1165–1173
8. 8Ziegler, Pedersen, Mowinckel & Biele (2016), PMID 27608958
9. 9Arnsten, stimulant effects on prefrontal signal-to-noise (e.g., 2009)
10. 10Bruton et al. (2025), Psychophysiology 62(2): e14750
11. 11Kutscheidt et al. (2019), ADHD Attention Deficit and Hyperactivity Disorders 11(4): 395–401
12. 12Wiersema & Godefroid (2018), PLOS ONE
13. 13Göz Tebrizcik et al. (2025), Biological Psychology 202: 109161
14. 14Zamariola et al. (2018), Biological Psychology 137: 12–17
15. 15Interoceptive performance and mental health symptoms (2026), Nature Mental Health
16. 16Heartbeat-evoked potentials and ADHD symptoms in adolescents (PubMed 37775030)
17. 17Panagiotidi, Overton & Stafford (2017), Acta Psychologica 181: 10–17
18. 18Schulze et al. (2022), European Archives of Psychiatry and Clinical Neuroscience
19. 19Audiovisual decision-making and impulsivity (2025), Scientific Reports
20. 20Ghanizadeh (2011), Psychiatry Investigation 8: 89–94; Bijlenga et al. (2017), European Psychiatry 43: 51–57
21. 21Barkley (1997), Psychological Bulletin 121(1): 65–94
22. 22Berk & Potts (1991), Journal of Abnormal Child Psychology 19(3): 357–377
23. 23Corkum et al. (2008), Contemporary Educational Psychology 33(1): 97–115
24. 24Winsler et al. (2000; 2007), private speech in children at risk for ADHD
25. 25Benfield et al. (2025), British Journal of Psychology
26. 26Alderson-Day & Fernyhough (2015), Psychological Bulletin 141(5): 931–965
27. 27Shaw et al. (2014), American Journal of Psychiatry 171(3): 276–293
28. 28Edel et al. (2010), European Journal of Medical Research 15(9): 403–409; Kiraz et al. (2021), Turkish Journal of Psychiatry 32(2): 109–117
29. 29Lieberman et al. (2007), Psychological Science 18(5): 421–428
30. 30Ra et al. (2018), JAMA 320(3): 255–263
31. 31Beyens, Valkenburg & Piotrowski (2018), PNAS 115(40): 9875–9881