The individual attribution of failure as the activating threat

Herman’s framework requires not merely powerlessness but powerlessness in the face of threat. In contexts of violence, the threat is physical and immediate. The specific, credible threat in the HDM context is the individual attribution of failure. It is certain in kind and unforeseeable in occasion: the architecture leaves nothing to attribute failure to except the person nearest it, and supplies nothing by which they could know whether or when it will come. What it costs runs a range of its own, from being quietly passed over — not consulted, not put forward, nothing said — to the loss of the position itself.

The team leader is still standing next to a missed deadline with nothing on record. What follows is not an argument about whether the date was reachable; nothing exists that could start one. Whether blame comes rests with the people above them, and nothing on record bears on it either way. If it does come, spoken or otherwise, there is nothing available to answer it with.

What a person will be held to is found out when they are held to it. The same absence does both jobs at once: it forecloses the response, and it leaves nothing by which the threat could be anticipated. Robert describes the result as a searchlight: ignored for months, then singled out over a complaint, a missed target, or nothing in particular, then dropped again as though it had never happened. Keeping out of its way was not available either: he describes being punished for doing something without permission he had not known he needed. What it leaves between the episodes is what he calls a constant, underlying stress — everyone on edge, wondering when they would next be caught.

That is not a milder version of a threat which arrives on schedule. De Berker, Rutledge, Mathys, Marshall, Cross, Dolan, and Bestmann (2016) found that subjective stress, pupil diameter and skin conductance all tracked irreducible uncertainty about whether an aversive event would occur, and that a small chance of it produced more stress than certainty of it. Watching for something that may or may not come is its own exposure, and the vigilance it produces reads the situation accurately rather than saying anything about the person watching.

Forced internal attribution

Abramson, Seligman, and Teasdale (1978) established that depression results from internal, stable, global attributions for negative events. Their framework distinguishes universal helplessness, where no one can control this outcome, from personal helplessness, where I specifically cannot though others could. Only the latter damages self-esteem, because the attribution is internal.

What the HDM context creates is something this framework does not theorise: forced internal attribution. The failure has a structural cause. The person may hold authority over a team of their own; no position in the chain comes with any way to question the mandate handed down to it. Nobody has to account for that mandate, and their perception that it could not be met had nowhere to land. Where it is attributed, it is attributed to them personally regardless. Nothing requires anyone to say it. Where it is not said, it shows in how the person is treated: whether they are consulted or passed over, whether their standing has quietly changed. The structure produces this without anyone deciding to. They are caught between what they know to be true and what the organisation, without saying so, has concluded. Accept the attribution and absorb the damage to self-esteem. Reject it and face the consequence of challenging authority. The structure has closed both exits.

Robert asked why he was not being put forward for more demanding roles. He was told he lacked bottom. It is not a quality anyone can be shown to have, and nothing had required a criterion for the choice, so there was nothing for the answer to be measured against either. What arrived in place of a reason was a property of the person, which he could neither contest nor correct. He has since collected the wider set of things people are told about themselves: that they lack resilience, focus, assertiveness or the knack of managing upwards, that they are too emotional, and that they do not speak up. The set does not have to cohere, which is what keeps it always available.

The attribution does not stay outside the person. Robert, looking back on this decades afterwards, describes what keeps people from speaking about it at all. We are not supposed to admit we are struggling, and there is a great deal of shame, guilt and embarrassment tied up with it: we think we have failed, we think we are inadequate, we think it is our own fault. That is forced internal attribution reported from inside, and the silence it produces is what keeps the pattern from ever being counted.

Shame is the specific response the experimental work associates with this class of threat. Gruenewald, Kemeny, Aziz, and Fahey (2004) had eighty-one participants perform identical speech and arithmetic tasks with an evaluating audience present or absent. Those evaluated showed greater increases in shame and greater losses of social self-esteem, while anxiety and performance self-esteem did not differ between the conditions. Salivary cortisol rose only where evaluation was present, and rose most in those whose shame rose most. What Robert reports is not a reaction peculiar to him. It is what a threat to standing produces in a person with no way to contest the judgement.

Same person, same outcome, two architectures

The clearest way to see what the architecture does is to hold everything else constant.

Where binding choices have to be substantiated, someone who runs the process to completion and flags a limit their perceptiveness cannot close has produced something the organisation holds: a named, identified risk. When the named risk materialises and the outcome fails, the architecture is doing what it should. The gap was mapped, and what could not be closed was declared in advance. Nobody is at fault, because the reasoning is on the record and the limit was stated before the work began.

Under HDM the same person, in the same situation, with the same perceptiveness, produces nothing the organisation holds. There was no occasion on which the limit could be stated. When the outcome fails, it falls on whoever stood nearest to it.

Same person, same perceptiveness, same outcome. Nothing about the individual explains the difference.

The set-up-to-fail mechanism is architectural, not interpersonal

Manzoni and Barsoux (1998) documented the Set-Up-to-Fail Syndrome at the interpersonal level: managers attribute subordinate failure internally, increase monitoring, which undermines autonomy, which degrades performance, which confirms the original attribution, a self-fulfilling prophecy operating within a specific boss-subordinate relationship.

The DFS analysis extends this to the structural level. The mechanism does not require a biased manager or a specific interpersonal dynamic. It requires only a decision-making architecture that combines directive authority with no substantiation requirement and no structural mechanism for challenge. The dynamic then operates independently of any individual relationship. And because it is architectural, it cannot be addressed by replacing individuals, improving relationships, or providing better management training.

Nothing is learned in learned helplessness

The objection to all of this is that something does get into the person: that people under these conditions acquire a passivity and carry it with them. Learned helplessness is the name that objection travels under, and it rests on an interpretation its own authors have since reversed.

Seligman and Maier (1967) found that animals given inescapable shock later failed to escape when escape became available, and called it learned helplessness. They reversed the interpretation half a century later (Maier and Seligman, 2016). The intervening neuroscience showed that passivity and heightened anxiety under prolonged aversive stimulation is the default response, mediated by serotonergic activity in the dorsal raphe nucleus. What has to be acquired is the opposite: the detection that the aversive condition can be altered by something the animal does. That detection activates the ventromedial prefrontal cortex, which then inhibits the default. Their own account is that the original theory had the direction wrong.

If passivity is the default and detection of control is what must be acquired, then HDM installs no disposition. It withholds the occasions from which one would be built. What would have to be detected is that the outcome a person is answerable for can be altered by what they perceive. Under HDM no occasion arises on which that happens, so the default is never displaced.

The conditions hopelessness theory identifies

Abramson, Metalsky, and Alloy (1989) extended the learned helplessness framework through hopelessness theory. Hopelessness arises not only from causal attributions but from inferred consequences and inferred characteristics about the self. The first is what will follow from the failure; the second is what this failure says about me as a person. In the HDM context, the inferred consequences can be severe: career damage, lowered status, reputational harm, potential termination. The self-implications are identity-threatening: years of professional expertise apparently not enough to stop a failure that same expertise saw coming.

The HDM context, as this paper reads it, generates the full set of conditions hopelessness theory identifies: internal attribution, stable attribution, global attribution, severe inferred consequences, and identity-threatening self-implications. What the theory predicts under these conditions is hopelessness. In the theory, hopelessness is the immediate cause of a form of depression that has symptoms of its own. Liu, Kleiman, Nestor, and Cheek (2015) reported in a twenty-five-year review that this theoretical account is robustly supported across populations.

No hostile act is required to develop Complex PTSD

A threat that nobody has to intend is still a threat. Sustained long enough, with the response foreclosed and the exit priced, it reaches the far end of the range without anybody doing anything to anybody.

The occupational health literature explains psychological harm by what the work does to people, not by what anyone is like. The World Health Organization defines burn-out as a syndrome resulting from chronic workplace stress that has not been successfully managed, and names nobody in it. An actor appears at one point on the spectrum and one only: in the exposures normally given as examples of the ICD-11 criterion of Complex PTSD. When it comes to torture, slavery, genocide campaigns, prolonged domestic violence and repeated childhood abuse somebody is doing something to somebody.

DFS argues that the range of psychological harm as found in organisations, from stress to Complex PTSD, is truly continuous also in the absence of hostile acts. The clinical literature already treats burnout, depression and the anxiety disorders as poorly separated from one another (Bianchi et al., 2015; Kotov et al., 2017). An actor present at one end and absent along the whole of the rest is a reflection of the material the criterion was built from, and not of what produces it. Trauma research studies people who present for treatment, and people present after something identifiable happened to them. The criterion’s own drafters describe the stressor as a gate criterion, state that trauma history is not determinative for the diagnosis, and locate what drives it in the symptom profile (Cloitre et al., 2013; Maercker et al., 2013).

Where an actor is present and deliberate, the acts are not what predicts the outcome. Basoglu, Livanou, and Crnobaric (2007) assessed 279 survivors of torture from the former Yugoslavia against a clinical PTSD interview. Physical torture did not significantly predict either PTSD or depression. Perceived distress and uncontrollability did. Sixty per cent of those who had undergone no physical torture at all met criteria for lifetime PTSD.

Where there is no actor at all, the same response is still produced. Subjective stress, pupil diameter and skin conductance track uncertainty about whether a shock will arrive, where the source of the stress is a probability (De Berker et al., 2016). Shame, lost social self-esteem and a cortisol rise follow from performing in front of an audience that can judge, where the audience does nothing hostile (Gruenewald et al., 2004). Passivity under prolonged aversive stimulation is the organism’s default state, and the apparatus delivering the stimulation has no intentions (Maier and Seligman, 2016). Across 208 laboratory studies, the largest cortisol response and the slowest recovery come from tasks that are uncontrollable and open to negative judgement (Dickerson and Kemeny, 2004). Persistent perceived job insecurity predicts worse health, while the job losses themselves do not account for the effect (Burgard et al., 2009). What produces the response in all five is uncontrollability, unpredictability and exposure to judgement. None of the three requires that anybody want anything.

Herman’s own account of what does the damage says the same. The victim is rendered helpless not by the nature of the threat but by the foreclosure of response, and helplessness is the essential insult. One mechanism runs the whole range, and no point on it requires anybody to be doing anything to anyone. HDM supplies the foreclosure of response, structurally, to everyone under it; the threat and the price of leaving come with the employment relationship. What varies is how much of it a person gets. Most people under these conditions will not reach the clinical end. What the evidence supports is that reaching it does not require a hostile act.

What this does not have behind it is an outcome study. Every case at the clinical end that has been measured had somebody identifiable in it, which is a fact about how cases are found rather than about what produced them, and is no evidence for the structural case either. Where nothing was required, nothing happens, so nothing is recorded, and a cohort cannot be assembled out of records of an absence. The loop closes here by mechanism and not by measurement. What would settle it exists: organisations at the floor of the abusive supervision instruments should still show the mechanism, and the International Trauma Questionnaire has not been run in a workplace sample with this architecture measured alongside it.