Volume V of the Middle Way Philosophy Series, to be published by Equinox June 2026

Human judgement is constantly distorted by biases that are due to the biological, social and cultural conditions of our lives. Cognitive psychology has identified a large number of these, whilst the critical thinking tradition has also found closely related fallacies. However, our response to these biases and fallacies is rarely considered as part of a whole personal practice. This book explores how the Middle Way can help us find the best practical response to our biases: neither believing our biased judgements uncritically, nor reacting against them to dismiss all judgement. The underlying problem of biased judgement is argued not to be one of ‘irrationality’, but of conflict created by absolutization (assuming we have the whole story). This conflict can be addressed effectively through Middle Way practices, as introduced in the earlier volumes of this series. The book offers a helpful overview and categorization of a wide range of biases and fallacies, seeking to identify balanced approaches to all of them in relation to each other.

Summary by Chapter

Introduction

Bias is an unavoidable embodied human state, but applying the Middle Way to bias helps us avoid both passive determinism and idealized ‘rationality’ in relation to them. Bias is meaning, but meaning does not predetermine judgement or belief, so biased judgement is the issue, not bias itself. Working to avoid biased judgement requires an understanding of biases in relation to each other, and an understanding of integrative practice in response to them.

1. Bias and Absolutization

1.a         Bias and biased judgement

The distinction between bias and biased judgement is that between meaning and belief. This distinction follows from embodied meaning, and the arguments for making it are practical ones: avoiding both the idealization of freewill and the paralysis of determinism, but instead connecting with our experience of responsibility.

1.b         Confirmation and its defence

Confirmation Bias is the product of a basic efficiency in organisms: maintaining either existing beliefs, or at least meaning that is pre-formatted for the easy creation of slightly modified beliefs. All biases consist in this ready-shaped meaning that can then be rigidified through biased judgement. Confirmation Bias also consists in a tendency to Attribute Substitution for ease, Belief Bias in reasoning, and Belief Disconfirmation Paradigm to avoid challenging alternative options.

1.c         Fallacy and bias

Fallacies have been traditionally presented as faults in logic, but this fails to capture the problem with them, which is unnecessary conflict created by failure to respect contextual rules of discussion. Biases and fallacies have traditionally focused on different conditions, but both identify the same practical problem of biased judgement. Biased judgement can often indeed be pursued with rigorous logic, dependent on isolated premises.

1.d         Absolutizing biased judgement

Biased judgement is absolutization, either in support of a bias or in reaction to it. Biased judgement follows all the features of absolutization detailed in volume I: mental proliferation, absoluteness of negations, restriction of options, reinforcing feedback loops, assumed system independence, fragility, representationalism, denial of embodiment, discontinuity, metaphysical defence, repression, conflict, projection, confirmation bias, substitution, and group binding.

1.e         Absolutizing reactions against bias

A reaction against bias that maintains the same framing is just as absolutizing as a biased judgement deceived positively by the bias, and can have all the same absolutizing features. Developmentally, however, this ‘flip’ can help us to move beyond the biased framing, as long as we retain awareness of avoiding the positive bias too, and are motivated to find third alternatives.

1.f          Embodied meaning, bias, and provisionality

Biases are habits of associative meaning that may be stronger or weaker. This enables us to work with biases through provisionality, which offers alternative meaning options in our judgement as we form beliefs. Provisionality involves slowness, suppression, synthesis, probabilizing and weighing up. To avoid biased judgement, we need to integrate meaning and desire through practices at those levels, rather than only belief.

1.g         Faith, confidence, and doubt

Faith is the more emotive and intuitive aspect of belief, and is more focused on its source than its propositional expression. This is trust when focused on a person. The positive, embodied faith we need to have in our beliefs can be called confidence and opposed to conflictual doubt. Faith and confidence can be absolutized in over-confidence, but when balanced are an important aspect of our capacity to avoid fragile biased belief.

1.h         Group authority and bias

Bias is group-oriented, and biased judgement is a shortcut for group binding, even for an individualistic counter-group. The established biases of Ingroup Bias, Social Proof, Groupthink and False Consensus support other biases, and are culturally rigidified by metaphysics. The ideological level of psychological development can provide a level of reflection and progressive values that can initially help to challenge group biases, aided by democracy and solitary reflection.

1.i           Absolutizing sources

We attempt to justify biased judgements either by circularity or by appealing to an absolutized source. The Genetic Fallacy is that our claim is true or false (not just incrementally credible) due solely to its source (which could be textual or personal). Experts are fallible, but should not be dismissed just because of their fallibility. We can overcome the Authority Bias behind this through using credibility assessment and through developing confidence.

1.j           Idealized models and metaphysical dogmas

Models give us consistent frames for possible beliefs, built up from metaphor, but the idealization of these models as final creates metaphysical dogma. Metaphysical beliefs fix and propagate our biased judgements by assuming that their anchors and frames are ‘true’ ones. Their representationalist assumptions also deny the meaningfulness of what lies beyond the frame.

2.   Biases of meaning

2.a         Anchoring and Availability

Anchoring is a general bias of meaning whereby associations from basic-level categories and embodied schemas limit our starting point for judgement. Availability Bias also explains our difficulty with finding new anchors, because we cannot recognize absences, our construction of meaning from unrelated phenomena, or our variations in emotional state. However, anchoring and availability limitations can be compensated for by adjusting away from our starting point, and through practices that contextualize emotional responses.

2.b         Framing

Framing limits the bounds of models or sets of possible beliefs in meaning, as distinct from Anchoring which is concerned with starting points for meaning. Our bias is to judge frames in isolation from each other, dependent on the assumptions attending concepts or specific descriptions. We cannot escape frames, but we can avoid absolutizing by widening them, reframing to synthesize previously separate frames.

2.c         Repression and manipulation

We can manipulate others by repressing other anchors and frames whilst promoting our favoured ones. Repression, which is also used internally, tries to eliminate alternative desires and beliefs and block awareness of them. We are vulnerable to such manipulation due to biases that can direct judgement through merely repeated meaning, as long as we fail to distinguish meaning from belief. Our motivation to repress is also due to biased judgement, and can be addressed by integration.

2.d         False Dichotomy

False Dichotomy is a result of framing of options into our favoured option and its negation, with any third repressed. It is the finality of this framing as ‘reality’, rather than the practical use of distinctions, that makes the dichotomy ‘false’. There is thus no contrast with ‘true dichotomies’, claimed examples of which turn out to be a priori distinctions from models. Philosophical dualism, craving and hatred can all further create conditions for dichotomizing.

2.e         Optimism and pessimism

Deeper forms of optimism and pessimism may arise from premature or delayed differentiation due to insecure attachment in infancy, perhaps exacerbated by later over-emphasis on left-hemisphere functions. The left hemisphere defaults to optimism, but can also appropriate pessimism from right hemisphere warnings, which helps to explain widespread flips between Negativity and Optimism Biases. They offer biased judgements that are distinct from positive or negative practical assessments.

2.f          Association Fallacies and Sympathetic Magic

Association fallacies arise from a failure to distinguish associative meaning (which is important in all sorts of ways) from justified belief. The Endowment Effect or its reversal takes association to be a basis of value, whilst Sympathetic Magic takes similarity or contagion to be a source of justifiable causal beliefs. In all of these we should not minimize the vital effects of meaning through association, but include these in a wider picture of judgement.

2.g         Ambiguity Aversion

Ambiguity and vagueness are unavoidable aspects of embodied meaning, but aversion to them creates bias because of the assumption of representationalism. This assumes ‘literalness’ as default, interpreting ambiguous experience as unambiguous. It becomes biased judgement by insisting deludedly on complete linguistic determination, or reacts by assuming an absence of meaning due to ambiguity. Positive acceptance of ambiguity and vagueness enables the appreciation of beauty.

2.h         Fake Precision and Absolutized Definitions

Ambiguity Aversion can result in Fake Precision, which assumes that precisely accurate language captures reality. A Slippery Slope Fallacy assumes that breaches of absolutized boundaries must lead to bad results. To defend Fake Precision, Essentializing of definitions with appeal to an absolutized source may be used. The Redefinition, Etymological or Original Language Fallacies offer further specific biased judgements that maintain Fake Precision.

2.i           Equivocation and Ad Hoc Thinking

Equivocation involves biased judgement when we fail to notice ways in which meaning has shifted within the practical terms of a given argument. Sometimes this is trivial or humorous, but is more serious when we switch between absolute and non-absolute uses of a term without recognition. Ad Hoc Argument and Motte and Bailey Argument are different methods of entrenching and defending equivocal arguments that are subject to this damaging confusion.

2.j           Metaphor and Weak Analogies

In embodied meaning, there is no ‘literal’ meaning to contrast with metaphor, so we need to assess the biases involved in using Weak Analogies differently. Rather than biased judgement taking a Weak Analogy as though it was literal, we need to see it as absolutizing one metaphor or another to prevent the integration of their perspectives. This occurs when we draw beliefs from the comparison that either take the analogy uncritically or react against it.

3.   Biases of subject and object

3.a         Projection and Overconfidence

Projection is a general feature of biased judgement whereby we assume all the features we experience are part of an object. Projection is supported by Over-confidence about the accuracy of our judgements, which influences both absolute judgements and probabilities. Over-confidence also extends to value assertions about others sharing our view, as well as over-estimations of logicality, depth of understanding, planning and skill. It can also flip into equally fragile Under-confidence.

3.b         Generalization and Straw Man

Generalizations are a normal and ‘reasonable’ part of our lives, but become biased when created hastily or applied sweepingly. They can also be refused altogether in reaction. Biased generalization can be avoided by provisionality and its practice. The same issues apply to one person in the Straw Man fallacy, which misrepresents individuals in argument by appropriating or lumping their position into a wider generalized one. Sensitivity to others’ distinctions, and ‘steel-manning’ can help avoid this.

3.c         Categorization biases

Categorization biases consist in selectivity of category (‘taking the best heuristic’) on scales of specificity vs generality, conceptual analysis vs synthesis, and abstract vs concrete. Our judgement is biased when we take these to offer final descriptions of objects. Disputes about how to interpret supervenience between levels of description are due to Fallacies of Composition and Division, whilst unhelpful synthesis and over-specialized analysis exhibit Jingle and Jangle Fallacies.

3.d         Appeals to Ignorance

Though practical conclusions derived from the absence of evidence may be justified in restricted conditions, Appeals to Ignorance involve a more general biased judgement that definite conclusions can be drawn from such absence. Appeals to Ignorance are rife in metaphysical philosophy, theology, and mysticism, and require specifically critical awareness in response.

3.e         Causal biases

Causal biases arise from our practical need to track linear causal relationships in a given context, which creates biased judgement when we neglect wider systemic relationships. We can absolutize cause on the evidence of correlation, fix causal direction, assume necessity or sufficiency, or allow specific outcomes to fix our belief. Determinism and indeterminism are culturally entrenched idealized models fixing linear causality, probably extended from early experience of agency.

3.f          Information Bias and Nirvana Fallacy

Information Bias is the assumption that we can always necessarily use more information, despite our embodied limitations in being able to use that information beyond a certain point. As a biased judgement this is an instance of the infinite rationalization of an absolute position. This can also create the Nirvana Fallacy when the information is about solutions to a problem, and imperfect solutions are constantly rejected due to their imperfection alone.

3.g         Views of other people

Recognizing other people as like us depends on right hemisphere functions, that can flip into instrumentalizing left hemisphere ones. These biases create biased judgement when we get stuck either in an initial negative objectification of someone (e.g in Ad Hominem), or in a rigidification of initial sympathy (e.g. in the Halo Effect). Stories motivated by archetypal functions can also lead us to project either positive or negative features on others.

3.h         Views of ourselves

Our introspective view of ourselves needs to be distinguished from the experiential framework of experience, and is subject to all the same biases as external objects, without being necessarily prior to those other biases. Our contingent inner experiences are rigidified into a fixed subject according to the Introspection Illusion, even though they give no grounds for either believing or disbelieving in a fixed self. Our identifications need to be stretched rather than our egoic ‘self’ destroyed.

3.i           How we think others see us

When we start using language to refer to ourselves, we start becoming dependent on others’ view of us. However, this can be biased by social needs, individual security and cultural influence, and absolutized by stress responses. We can overestimate others’ attention to us and understanding of us. Others’ expectations can shape us and ambiguous terms be interpreted to refer to us. Such sensitivity to feedback can serve a helpful role, but needs to be incrementalized and contextualized.

4.   Biases of space and time

4.a         Domain dependence

Embodiment makes us subject to limitations in space through anchoring and availability, but we absolutize those limitations either by fixed boundaries or by idealized universality. Domain Dependence can unnecessarily limit attention, meaning, functional use of objects, and sampling as a basis of judgement. The limiting domain can be social or conceptual, as well as temporal, in extension of our experience of space. We overcome such limitations through stretch.

4.b         Getting stuck in the past

We get stuck in the past, like other temporal biases, due to discontinuity in the left hemisphere experience of time, cutting off wider present awareness. Thus we idealize the past, identify with its investments, fail to change habits from it, or have a decontextualized view of it. Even in the short-term we recall events that fit our goal-orientation rather than whole processes. Studying the past reflectively may help us stretch these distortions and avoid absolutizing them.

4.c         Forgetting the Past

The concerns of the present also tend to crowd out helpful awareness of the past, absolutized when we take the present as the whole story. This Appeal to the Present can be seen when we judge the past only on what has survived, when we falsify the past to fit the present, when we are either obsessed with or neglectful of current news, and when we hold current goals in mind at the expense of wider awareness. Practice can help us be aware of the past, as well as ‘living in the present’.

4.d         Discounting the future

We discount the future when we allow present concerns to eclipse it, despite the strength of evidence for future events. Such discounting can be built into a ‘rational’ economic system as well as being due to individual conflict, creating procrastination in our personal activities, and unawareness of future emotional states or changes in character. We’re resistant to changing the status quo because it costs more energy.

4.e         Idealizing the future

Biased judgements about the future are isolated from the context offered by the present, projecting certainties regardless of the evidence. Optimistic bias is most common due to left hemisphere dominance, whether involving Impatience, Neomania, or Forecast Illusion. Most forecasts prove to be wrong and are the product of Over-Confidence. Pessimistic bias, from a left hemisphere capture of right hemisphere insights, creates catastrophizing, Fear of Regret, or Appeals to Fear.

4.f          Errors in probabilizing and probability

Our judgement of probabilities is vital for (but not exclusive to) our response to future events, and does not depend on precise calculation. It’s more important to avoid biased judgements of total certainty or dismissal. When probabilizing we also need to avoid over-focus on either normal or unusual events, expecting causal patterns in evidently random events, assuming the conditions of engineered randomness in a complex system, or a focus on narrower categories than the probability applies to.

5.   Biases of value

5.a         Value biases in general

Values are not entirely distinct from facts, but are more focused on sustainable motives. Our understanding of values depends on models: either aesthetic, socio-political, or ethical, with ethical models attempting some integration of the socio-political values identified by Haidt. Our value biases consist in the use of some models, better adapted to certain conditions, to the exclusion of others, with biased judgement giving absolute justification to this exclusivity.

5.b         Aesthetic biases

Aesthetic biases are narrow interpretations of our experience of beauty from only one point of view. We may assume that beauty is in the object itself, that it is ‘merely subjective’, that it can be understood entirely in terms of pleasure, that it can be reduced to an instrumental function of some kind, or that it can be reduced to a conceptual analysis. All of these perspectives can contribute to our understanding of beauty, but become biased beliefs when they detract from our attention to it.

5.c         Biases of desire and pleasure

Biases of desire give value to short-term desires over long-term ones, also identifying them with pleasure. They include various forms of Wishful Thinking, where other assumptions are distorted, and the Hedonic Treadmill, where the pursuit of pleasure undermines its achievement. In reaction, we can flip to Asceticism in which we overestimate our ‘willpower’, or utilitarianism where we try to identify with ‘rationally’ defined aggregate pleasure. An integration model is needed to avoid these.

5.d         Authority, rules, and freedom

The value bias towards authority develops from parental relationships in infancy, whence we learn suppression as well as repression. This authority can be transferred to propositional rules as Legalism, with biased judgement developing as we absolutize either the authority or their rules. Deontological ethics and the value of freedom can each help to contextualize Legalism, but Antinomianism and Libertarianism can absolutize the reaction against rules and authorities.

5.e         Justice: absolute and relative

The value of justice of all kinds develops from our experience of similarity of treatment with siblings or peers, becoming biased judgement when it excludes other values. This absolutization can take the form of projection of the administration of justice (Cosmic Justice) or at least its meaning (Moral Naturalism) onto the universe. In reaction, it can also create the Relativist Fallacy of assuming justice is inherent to a given domain, rather than a value applied in our experience through judgement.

6.   Biases of responsibility

6.a         Total and Zero Responsibility Biases

Responsibility is relevant to all judgements, but is particularly concerned in our assumptions when individuals have applied values. Our dichotomous bias when attributing responsibility or its absence is created by freewill and determinism assumptions and Legalism, resulting in the Total and Zero Responsibility Biases. Biased judgement assumes this dichotomy to be necessary. All the responsibility biases in this section are applications of this dichotomy.

6.b         Actor/ Observer Biases

Actor/observer biases emerge from the differing perspectives of acting for oneself and observing others act, but become biased judgement when we apply the Total v Zero Responsibility dichotomy. They lead us to attribute responsibility to good judgements we identify with, but not for ones we do not identify with, and to deny responsibility for bad judgements we identify with, but not for those we do not identify with. Reflection on the similarities between all perspectives can modify this bias.

6.c         Action and omission

The Action and Omission Biases are an opposed pair that lead us to act when non-action would be more appropriate, or fail to act when action would be more appropriate. We fear being held responsible if we fail to visibly act, or we try to exploit people’s relative lack of attention to omission. The distinction between acts and omissions, however, is an arbitrary matter of description, and relaxation of our expectations can avoid biased judgement.

6.d         Responsibility and time

Whether an event is past or future is a further source of Total or Zero Responsibility Biases, with past events not just unchangeable, but it being assumed they could not have been different. Assumptions of past freewill, however, can also be fixed, and the future also idealized as entirely free in contrast to the determined past.

6.e         Responsibility and effort

Although effort is an important and normal part of the way we develop responsibility, we can also make biased judgements by assuming total responsibility for our projects and thus not letting go of them. Ambiguous contributions to social actions, though, may lead us to assume no responsibility, but instead make a token effort to satisfy others’ expectations. The Buddhist concept of ‘balanced effort’ can also help us find appropriate levels of responsibility.

6.f          The formalism of choice

The Total Responsibility Fallacy can also occur when we identify responsibility with ‘choice’, but when choice may be merely formalistic and without the meaning required for optionality. This can be the case when we have too many choices, lack the energy to choose meaningfully, or are resistant to choosing because of the loss of future choice. Responsibility claimed as the product of choice rather than optionality can also be formalistically denied as Bad Faith.

Conclusion: the critical practice of contextualization

This book has applied the Middle Way to overcoming biased judgement by showing it as part of a wider practice of contextualization. The six elements of the cross-disciplinary skill of critical thinking (reasoning, justification, recognizing contexts, avoiding biases and fallscies, interpretation, and credibility assessment) all work by contextualizing through the direction of attention to new meaning. Other practices also contribute to our long-term addressing of bias, including mindfulness, the arts, and general education.