A. The Need to Eliminate Vagueness

The drafting of laws is an integral part of any legislative process. While the nuances of drafting vary across jurisdictions, certain trends apply almost universally irrespective of jurisdiction. The need to eliminate vagueness from statutory language is one such trend.

In the United Kingdom, the Drafting Guidance to the Office of the Parliamentary Counsel requires the use of precise and concrete words, as opposed to vague and abstract ones.^{Footnote 8} Similarly, the Guide to Making Federal Acts and Regulations in Canada states that “perhaps the most important function of legislative drafting conventions is to reduce the ambiguity or vagueness of a natural language such as English or French”.^{Footnote 9} In a similar vein, the Legislative Department of the Indian government outlines their vision and mission through a list of goals that should be kept in mind when drafting all kinds of legislations. Of these goals, the goal of precision is given very high importance.^{Footnote 10} A similar approach is taken in the United States of America,^{Footnote 11} Australia^{Footnote 12} and in parts of Europe.^{Footnote 13}

This general trend against vagueness affects the way laws are drafted. For instance, drafters are often encouraged to eliminate vague words as they can be the source of frequent litigation.^{Footnote 14} Examples of these vague words include certain commonly used phrases: “best efforts”, “fair”, “reasonable”, “as it may deem fit” and other such phrases.^{Footnote 15} Drafters are instructed to either replace these words with precise words^{Footnote 16} or define these words in the definition sections of statutes with precise words.^{Footnote 17} These instructions seek to increase objectivity in legal language at the stage of drafting and prevent any significant subjective deliberation at the adjudicatory stage.

These practices and drafting guidelines are supported by a range of academic texts across jurisdictions that encourage drafters to follow the maxim: precise words are always to be preferred over vague ones.^{Footnote 18} In fact, this maxim is so intuitive that most of us rarely question it. The intuitive function of this maxim coupled with its universality shows that the maxim is entrenched in our understanding of law itself.

B. The Desideratum of Clarity

As we have seen, drafters are encouraged to eliminate vagueness to ensure that laws are as clear as possible. Interestingly, in most of the drafting guidelines referred to earlier, the words “clear” and “precise” are used synonymously.^{Footnote 19} However, clarity and precision are distinct concepts. An understanding of this distinction is crucial to help understand our reservations, if any, against vague expressions.

The notion of clarity is perhaps best understood by referring to Lon Fuller’s detailed discussion of this topic in his 1964 book, ‘The Morality of Law’.^{Footnote 20} In a lesser known part of this widely read book, Fuller makes important observations on clarity. If read closely, Fuller himself advances a subtle distinction between the concept of clarity and the concept of precision in this text—one which will inform the analysis made in the latter portions of this paper.

The purpose of law is to “subject human conduct to the governance of rules”, writes Fuller.^{Footnote 21} A lawmaker seeks to govern human conduct by enacting rules. In describing the existence conditions for a legal system, Fuller states that wherever law exists lawmakers must have complied with the formal constraints on law-making, to a certain extent. He proceeds to outline what these formal constraints are and arrives at eight desiderata which are often referred to as the ‘principles of legality’.

Fuller arrives at the principles of legality through a parable about an imaginary ruler, Rex.^{Footnote 22} Rex tries to enact rules that effectively guide human behaviour. Rex fails to do so on eight counts. Discussing each of Rex’s failures, Fuller outlines eight principles of legality. These are as follows: (i) laws must consist of general rules; (ii) laws must be publicised; (iii) retroactive legislation must be used within adequate bounds; (iv) laws must be as clear as possible; (v) laws must be non-contradictory (it must not be the case that one law can only be obeyed by violating another); (vi) laws must not demand the fulfilment of impossible actions by the subjects; (vii) laws must possess sufficient stability over time; and (viii) laws must be congruent—“officials must act and decide on the basis of and in accordance with the law”.^{Footnote 23} A total failure to comply with any one of these formal principles, results in something that cannot be properly called a system of law.^{Footnote 24} Therefore, there is a necessary connection between a certain degree of compliance with the principles of legality and the existence of law.

Fuller’s list of eight desiderata is a surprisingly comprehensive set of existence conditions for a legal system. Even Fuller’s theoretical rivals, the legal positivists, endorse Fuller’s list. What legal positivists take issue with is Fuller’s claim that these formal principles amount to an internal morality of law.^{Footnote 25} This contentious claim will not be discussed in this paper. Instead, I focus on desideratum (iv) in Fuller’s list—the desideratum of clarity. What does Fuller mean when he says that laws must be as clear as possible?

Fuller notes that the desideratum of clarity has been only marginally discussed in jurisprudence. He attributes this failure to the positivists who, owing to their reluctance to admit that an unclear law is less a law than a clear one, have only very briefly addressed the concept of clarity.^{Footnote 26} Whether or not Fuller is correct in attributing this failure to the positivists, he correctly points at the lack of literature on clarity—highlighting the scarce treatment given to the topic in the works of the likes of Bentham and Austin.^{Footnote 27}

It is in this contextual backdrop that Fuller’s discussion of clarity is important. Fuller criticises “obscure and incoherent” legislations, arguing that it is the legislature’s responsibility to ensure that such laws are not drafted.^{Footnote 28} However, in an often overlooked paragraph in his book, Fuller takes issue with what he calls “specious clarity”.^{Footnote 29} Drawing from Aristotle, he agrees that we cannot be more exact than the nature of the subject matter we are dealing with.^{Footnote 30} Through this statement, he concedes that there are cases in which we cannot be exact or precise in drafting laws. So, in these cases where exactness is not possible, a “specious clarity” is much more damaging than an “honest open-ended vagueness”.^{Footnote 31}

Fuller is thus distinguishing between two concepts: clarity and exactness or precision. While aspiring for the former prevents obscure and incoherent legislation, the latter is neither achievable nor desirable in certain cases. In cases where precision is impossible, Fuller admits that an “honest open-ended” vagueness can be of value. Contrary to what Fuller intended, our understandings of the rule of law often have very little room for this open-minded vagueness.

C. Clarity, Precision and Vagueness

A close reading of Fuller necessitates properly distinguishing between the concepts of clarity and precision. Fuller tells us that clarity is important to strive for. I agree. I will establish that clarity is correctly contrasted with ambiguity and precision is correctly contrasted with vagueness. It is important to distinguish between these concepts to understand what our goal is. If we must strive for clarity, our goal is to eliminate ambiguity. It is not to eliminate vagueness.

Ambiguity is different from vagueness. An ambiguous word has more than one meaning and creates confusion about which meaning is in use.^{Footnote 32} Philosophers of language uphold this definition of ambiguity and tell us that ambiguity can be of many types.^{Footnote 33} The simplest case, similar to what has just been discussed, is a case of lexical ambiguity – wherein a single term has two or more meanings.^{Footnote 34} Consider the word pupil. In one sense, the word pupil refers to a black hole located in the center of the iris of a human eye. In another sense, the word pupil refers to a student in an educational institution. If the word pupil is used without clarification and context, lexical ambiguity may arise regarding the use of the word.

There are several other types of ambiguity as well such as structural ambiguity, systematic ambiguity, type-token ambiguity and object-act ambiguity.^{Footnote 35} While a detailed discussion on ambiguity falls outside the scope of this paper, it is important to note that ambiguity leads to obscurity and a lack of clarity in a general sense. The receiver, listener or reader is unsure which meaning of the word is being used.

This ambiguity leads to a lack of clarity in the Fullerian sense—an obscurity through which the reader of a law does not know to what the law refers. What do the words in a statute mean, considering there are at least two meanings? Such ambiguity is to be avoided, Fuller tells us, to ensure that the goal of legislative clarity is achieved.

This is largely indisputable. We must know which meaning of a word is being referred to in a law to be able to follow it. And so, if we accept that we need to strive for clarity, we must seek to eliminate ambiguity.

While clarity should be aspired for and ambiguity should certainly be reduced, the same cannot be said for the precision-vagueness binary. Vagueness is different from ambiguity.^{Footnote 36}A vague word, unlike an ambiguous one, has only one meaning but its application is unclear. An expression is vague if there are borderline cases for its application.^{Footnote 37} When there are no borderline cases for the application of an expression, it is a precise expression.^{Footnote 38}

I will discuss the concept of borderline cases in the next part of this section. For now, I want to call attention to the differences between clarity and precision, and the distinction between ambiguity and vagueness. I agree with Fuller (as I suspect most legal theorists do): we must reduce ambiguity, as we have to strive for clarity. The same cannot be said for vagueness. I will argue that vagueness is both inevitable and sometimes desirable over precision in legal language. And so, vague words do not necessarily make for bad laws.

2. The Inevitability and Limited Desirability of Vagueness

In this section, I will draw from central debates in jurisprudence on the nature of language and its role in legal processes to argue that the desire for precision and the need to over-define legal terms is flawed. To do so, it is crucial to first frame the argument made by those in favour of eliminating vagueness. The argument against vagueness can be summarised using the three premises and the conclusion that follows:

While I agree with P1 (as I have shown vagueness is correctly understood as the opposite of precision), I take issue with P2 and P3 in this paper. If P2 and P3 are false, conclusion C does not follow. I hope that this analysis can conclusively demonstrate the flaws in the above-mentioned argument and caution against its use in policy, academia and the practice of legal drafting.

A. Inevitability

Legal philosophers are interested in studying the concept of law. Central to legal philosophy are questions on the relationship between law and language. I will briefly enter the jurisprudential debate on legal interpretation and language to demonstrate that a study of the philosophy of language leads us to the conclusion that vagueness is inevitable. Efforts to eliminate vagueness are redundant.

H.L.A. Hart’s classic jurisprudential text, ‘The Concept of Law’, is an effective entry point into the law-language discussion in the context of judicial interpretation. In Chapter 7, Hart discusses a middle-ground theory between the two prevalent approaches that he refers to as “formalism” and “rule-skepticism”.^{Footnote 39} Formalism can take two forms. In its descriptive form, formalism is the claim that all judicial decisions are clearly determined by pre-existing legal rules.^{Footnote 40} In its prescriptive form, formalism is the claim that all judicial decisions should be clearly determined by pre-existing legal rules.^{Footnote 41} On the other hand, rule sceptics argue that pre-existing legal rules almost never guide judicial making.^{Footnote 42}

Hart argues that while the formalists or textualists fail to account for the “open texture of language”, the rule sceptics exaggerate the reach of this very open texture.^{Footnote 43} Hart tries to forge a middle ground between these two theoretical positions, which he constructs as opposing extremes on a theoretical continuum.

For Hart, all words and general legal concepts have a core of settled meaning or certainty and a “penumbra” of uncertain application or doubt.^{Footnote 44} Judges must check if the case is a core case. This classification is not a purely logical deductive process. While it is obvious that some instances fall within the “core” of settled meaning, in other cases, the classification is uncertain.^{Footnote 45} In these penumbral cases, pre-existing rules cannot guide judicial decision-making.^{Footnote 46} Hart’s middle ground suggests that pre-existing legal rules clearly regulate judicial decisions, but only in core cases and not in penumbral or borderline cases. Judges must check if the case is a core case. If it is, they must apply the pre-existing legal rule. If not, the case falls outside the scope of the law. They must then apply their own discretion to decide the case at hand. Their decisions are no longer within the ambit of the pre-existing rule or the law.

Hart’s middle ground manages to capture a certain intuitive idea of judicial decision-making. Rules seem to clearly guide judicial verdicts, unless the words in themselves are vague. This construction of adjudication lends itself to the instructions given to legal drafters to eliminate vague words and reduce the possibility of borderline or penumbral cases.^{Footnote 47} The problem, however, is that while the jurisprudential conversation on this topic has advanced greatly from where Hart left us, the real-world approaches to policy and drafting seem to be still greatly informed by a Hartian understanding of the core and the penumbra. If understood through the Hartian core-penumbra frame, drafters seek to clearly outline the core and distinguish it from the penumbra, whilst simultaneously trying to eliminate or reduce the possibility of penumbral cases.

This approach to drafting statutes fails. Hart’s middle ground theory has come under attack, particularly for its failure to understand vagueness and the inevitable structural complexities of human language. Though Hart does briefly acknowledge that open-ended standards are sometimes effective, his theory is based on a construction of rules as precise maxims by which people can regulate their behaviour (even though in some exceptional cases vague standards may be adopted).^{Footnote 48} Hart’s theory is thus incomplete because he fails to account for the extent and prevalence of vagueness in legal rules. When we understand vagueness, we see that words often lack a clear core of settled meaning—and thus, Hart’s distinction of core and penumbral cases does not capture the complexities of vagueness.

Let us then briefly survey the recent jurisprudential conversation on vagueness. Philosophers of language have extensively discussed ‘vagueness’ or vague predicates. A vague predicate is defined by a lack of a sharp linguistic boundary.^{Footnote 49} A predicate A is vague if there are instances in which it is not determinate whether some object ‘x’ is A or not.^{Footnote 50} At what point of hair loss is a person deemed to be bald? We cannot clearly say. This is the philosophical problem of ‘vagueness’.

Consider n is the amount of hair on a person’s head. For any n, if x(n) is bald, then x(n+1) is bald. This would mean that if x(n+1) is bald, then x(n+2) is bald and so on. If we apply this reasoning, we are led to the (clearly absurd) conclusion that even a person with a full head of hair is bald.^{Footnote 51} This seems false, but we seem to have reached this conclusion by a sound chain of reasoning. Can the answer lie between n+1 and a man with a full hair of head? Timothy Endicott says that legal cultures adopt the belief that the addition or subtraction of one strand of hair will not affect the application of the word ‘bald’, though a sufficient change in the number of hairs will do so.^{Footnote 52} Endicott calls this the ‘tolerance principle’.^{Footnote 53} Where on this continuum does the answer lie, though? At what point of hair growth will a person lose their baldness?

Indeed, Hart’s failure can be most accurately understood by reference to the Sorites Paradox.^{Footnote 54} The paradox involves a heap of sand, from which grains are individually removed. Suppose we remove a single grain at a time, at what point does the heap change from a heap to a non-heap? There is no clear point we can outline. There is no determinable boundary between the core and the penumbra.

These issues capture the problem with Hart’s account. At what point does the core give way to the penumbra? We do not know. When we try to determine a clear boundary line for where the core ends and where the penumbra begins, we are engaged in an exercise that is both impossible and futile. The inevitability of vagueness in human language, no matter how well or precisely words are defined, is not something that drafters can escape.

Therefore, vagueness is inevitable in the structures of human language. Legal theorists have debated the extent of this vagueness—arguing about whether we can never know what particular instances a general word applies to (the ‘radical indeterminacy thesis’) or whether we often do not know this (the moderate thesis).^{Footnote 55} Whichever of these approaches we take, we must accept the fact that more often than not, vagueness is inevitable in the process of drafting legal rules. If we accept the inevitability of vagueness in the process of legal rulemaking, we can better deal with the challenges it poses for legal systems.

Therefore, we often cannot achieve precision. Efforts to achieve precision i. e. to clearly outline the boundary line between the core and penumbra are thus ineffective. As vagueness is inevitable in the languages that human beings use (as compared to a perfectly scientific-logical language),^{Footnote 56} P2 is disproved.

Having shown that P2 is false, we have reason enough not to strive for precision. It makes hardly any sense to aim for precision if this is an impossibility. However, I will go a step further to prove that vagueness is not only inevitable, but also, desirable. Vagueness can have value.

B. Desirability

By arguing that vagueness is inevitable, I have made a descriptive claim. I will now proceed to make a second, normative claim about vagueness: vagueness can have value. If this is shown, P3 cannot stand. Precision is not always desirable if vagueness sometimes is.

Contrary to the prevalent conception of vagueness as a problem in legal language, some legal philosophers have made a case in favour of vagueness. While Jeremy Waldron claims that it furthers people’s own ability for practical deliberation, Endicott argues that vagueness is a necessary route in certain legal spheres.^{Footnote 57} Endicott points to what he calls the ‘normative principle’—a norm must provide a reason for action, and a purpose.^{Footnote 58} A norm that does not abide by this principle is arbitrary and lacks a reasoned justification.^{Footnote 59} For any norm to be adopted, it must be arbitrary to the minimum and we must have good reasons for adopting it.

Using Endicott’s ‘normative principle’, let us discuss P3. According to P3, precision is always desirable. Is this norm reasonable? Or does it also give way to arbitrariness? Endicott argues that the need for precision can lead to arbitrary consequences, just like vagueness can.^{Footnote 60} Both precision and vagueness, thus, are arbitrary to different degrees in different contexts.

Consider the example of laws that provide a minimum age for alcohol consumption. In most jurisdictions, the law stipulates that a person must be above a certain age to be able to legally drink. This age is set at eighteen years of age in some jurisdictions, while others choose a lower or higher age limit. The reason behind age requirements for alcohol consumption is that due to alcohol’s intoxicating effects, a person must have reached an adequate stage of maturity to be able to drink and take responsibility for their actions. In jurisdictions where eighteen years is set as the age, certainly not everyone above the age of eighteen years can be said to possess the required level of maturity and not everyone of below eighteen years of age can be said to be lacking in this maturity. The precise age limit is thus arbitrary.

However, in a case like this, vagueness would also lead to arbitrariness. If there were a legal standard that permitted those ‘mature enough to handle alcohol’ to consume alcohol, the case-by-case determination of whether people had this maturity to drink or not would be left to those selling alcohol. This power could easily be abused. As Endicott notes, in cases like this, perhaps the “arbitrariness of the precise standard may itself have an important expressive function, signifying the community’s refusal to draw invidious distinctions among persons”.^{Footnote 61} In cases like this, both precision and vagueness are arbitrary—but the arbitrariness of the former is preferred because of the area of law and the context in which regulation occurs. Therefore, when regulating different forms of conduct, we must ask whether vagueness or precision are less arbitrary and regulate conduct accordingly.

As critics of vagueness argue, the arbitrariness of vagueness lies in the fact that it vests judges with what some regard as excessive judicial discretion. Their main problem is that the use of vague language allows for unguided judicial discretion. Endicott suggests using the ‘Similarity Model’ to tackle this issue.^{Footnote 62}According to this model, a term will apply to an object if the object resembles the paradigm sufficiently in relevant respects.^{Footnote 63} To obtain more clarity on how judges can apply the model, I suggest turning briefly to Ludwig Wittgenstein.^{Footnote 64} Wittgenstein’s theory of ‘family resemblances’ can explain vagueness in the following manner:

This account clearly captures the complex, contextual relationship that words share with meaning. The meaning of words is not settled in isolation of context, and is not a determinable core, as Hart suggests. In this scenario, then, determining whether or not a thing can be classified under a term requires a judge to interpret whether the word fits within a larger interpretative family or not. The question of whether a thing resembles a term in the relevant ways involves a complex adjudicatory deliberation. The arbitrariness of vagueness thus can be dealt with by developing techniques that judges can follow to adjudicate cases meaningfully and tackle the problem of vagueness.^{Footnote 66}

In conclusion, both precision and vagueness can be arbitrary. We need to weigh which form of arbitrariness is worse, and in what legal context. While precision is sometimes desirable, so is vagueness. Thus, as a norm, vagueness possesses ‘limited desirability’—it is desirable in some cases over precision, but not in others.

P3 states that precision is always desirable—a premise that, thus, collapses. While precision may sometimes be more desirable than vagueness, presenting precision as the default option and the aspirational goal for legal drafters is based on flawed theoretical assertions such as P3.

In the next two sections of this paper, I will examine this jurisprudential insight on vagueness through the lens of feminist theory. I hope that these reflections can centralise feminist theory in the current jurisprudential debates on law and language.

A. Relationality and Care

There are some areas of law where vagueness is more valuable than precision. In this paper, I suggest that vagueness is more valuable than precision in areas of law that regulate personal relationships or relationships of care. To do so, I draw from the feminist ethics of care—a frame, that though highly contested, remains one of the most influential ways of understanding relationships of care.

Care ethicists primarily derive their work from Carol Gilligan’s landmark book, ‘In a Different Voice’ which has had a profound impact on feminist theory across disciplines.^{Footnote 91} In her challenge to psychologist Lawrence Kohlberg’s theory, which glorified the rational, autonomous reasoner as the epitome of moral reasoning, Gilligan suggested that women grow to see individuals as shaped through connectedness, often linked through emotion and interdependent through care.^{Footnote 92} This relational approach to ethics is a structurally “different voice” from the individualistic approaches in canonical moral philosophy which construct individuals as detached and disembodied.^{Footnote 93} For care ethicists, this emphasis on relationality and subjectivity is an equally valid moral voice as compared to canonical moral theories that emphasise individualism and objectivity.

Care ethicists thus centralise particularity and context as important values within decision-making. They insist that crucial features of human life cannot, and should not, be judged without reference to context. Virginia Held describes care ethics as rooted within lived experiences, with an emphasis on embodiment and particularity.^{Footnote 94} For Held, we can only understand relationality, particularly intimate relationships in structures such as families, if we pay attention to the context of each relationship and the lived experience of the relational actors. Stated simply, each relationship is different. Care is complex and subjective. We cannot judge the nature and the quality of relationships through tools of abstraction.

This paradigm shift from the detached positivist approaches in legal philosophy is crucial for our discussion on vagueness. When legal drafters are told to be precise, they attempt to use words in a manner that pre-existing legal rules can completely determine judicial decisions.^{Footnote 95} But is this possible in areas of law that deal with intimate human relationships?

Consider the domain of family law. While we can share intimate relationships with persons outside heteronormative, patriarchal structures of families, family law continues to legally regulate a large part of intimate human relationships—or what are relationships of care. One of the most complex issues in family law are laws relating to child custody. Which parent should get custody of the child? Which parent is a “better” parent? Where will the child be given better and more care? A brief examination of these issues on custody reveals the complexity that care ethicists highlight.

The ‘best interest of the child’ threshold has been used almost universally to decide child custody disputes. Under the United Nations Convention on the Rights of the Child, “in all actions concerning children, whether undertaken by public or private social welfare institutions, courts of law, administrative authorities or legislative bodies, the best interests of the child shall be a primary consideration.”^{Footnote 96}

What is in the best interests of the child or what best furthers the child’s welfare?^{Footnote 97} Time and again, there have been efforts in different parts of the world to make this threshold more precise. American legal scholar, Robert H. Mnookin, and others in the United States of America have insisted that the ‘best interests’ threshold should be replaced with a more objective rule, as it currently provides no direction to judges.^{Footnote 98} In 2013, Carl Funderburk, a Special Judge in Oklahoma USA, wrote a controversial paper that argued that the best interest threshold should not be left vague especially in an area such as child custody.^{Footnote 99} The best interests threshold has also been challenged in the American Supreme Court and the court has heard petitions to declare the threshold ‘void for vagueness’.^{Footnote 100} According to a brief submitted to the Supreme Court in Troxel v. Granville, “The best interests test has long been the subject of academic as well as judicial criticism for being indeterminate, providing little guidance on how to weigh the different needs of individual children, especially as they change over time.” ^{Footnote 101}

These calls to precisely define what lies in a child’s best interests have been made in the United Kingdom^{Footnote 102} and in several Nordic countries as well.^{Footnote 103} Additionally, the Law Commission of India also suggested the need to discuss replacing the ‘best interest threshold’ with a more predictable rule-based law in its 257th Report.^{Footnote 104} These developments, and perhaps several others across the world, are supported by academic texts that have suggested making the threshold more precise, as vagueness is fundamentally antithetical to the rule of law even in an area like child custody.^{Footnote 105}

This position on child custody laws is both flawed and dangerous. First, the need for precision in child custody laws is reflective of a larger legal culture that prioritises precision over vagueness at all costs. As I have discussed at length in this paper, precision is neither achievable nor is it universally desirable. We need to prioritise either precision over vagueness or vice versa depending on the area of law we are regulating.

Second, the area of child custody laws deals with a network of intimate relationships—parent and child, children with one another, parent and parent. The web of relationships involved here can only be appreciated on a case-to-case basis. An approach that seeks to define what constitutes better parenting at the stage of drafting a law misses the point about relationality.

As care ethicists note, the practice, labour and emotive value of care is complex and subjective. These aspects of care are hard to put a measure on through precise legal language. When pushed towards precision, drafters resort to drafting laws using objective metrics such as, which parent is better off financially, or which parent lives closer to the child’s school, amongst other such factors.^{Footnote 106}

Such attempts at precision can create absurd judicial results. Often, the parent who qualifies for custody on an objective legal list, in fact, does not provide care—understood as labour and practice—towards their child.^{Footnote 107} If we aim for precision over vagueness, judges must apply the law as it is and grant custody to the parent who is perhaps better off financially, but who certainly “cares less”. This is the price we pay for precision. We make the wrong decisions, because we are so caught up in precisely defining rules through linguistic devices at the stage of drafting.

Drawing from the analysis in section 2, we know that no matter what we do, we cannot eliminate borderline cases or the possibility of vagueness in language. Instead, we must appreciate the role vagueness plays, as illustrated through our discussion of child custody laws. As care ethicists tell us, there is value in this subjectivity. For it is only through context and subjectivity that we understand relationships of care. The sooner we get comfortable with this subjectivity and vagueness, the better.

B. Intersectionality and Discrimination

Another area in which vagueness could be of more value than precision is discrimination law. The law on discrimination relates to incidents of unequal or unjust treatment based on a range of personal characteristics—a person’s gender, caste, race, disability, national origin, religion, sexual orientation etc. These identities constitute different social locations, which often combine together to create discrimination of a compounded nature. Such discrimination can be addressed through the theoretical lens of intersectionality.

Intersectionality, within feminist theory, has gained currency as an important theoretical and political tool towards social justice and transformation. Traced back to Kimberle Williams Crenshaw’s landmark work on the compounded discrimination faced by black women in America, intersectional theorists argue that individuals are made up of several social identities—gender, race, caste, class and other social identities.^{Footnote 108} These identities together determine the extent and nature of discrimination that people face.^{Footnote 109}

An understanding of discrimination thus requires an internalisation of the idea that there are several axes of identities under which an individual can face discrimination. The combinations of such axes are several, and indeterminate. This leads to a unique kind of complexity. Recognising this complexity, feminist theorists have emphasised lived experiences and contexts in determining whether or not discrimination exists, particularly of an intersectional kind.

As a result of this complexity of intersecting axes, legal systems can only account for the unique kind of discrimination faced by a person by understanding context and uncovering the identities which are interacting in a certain case to cause discrimination against an individual. In this case, then, vagueness is more suitable than precision as a regulatory frame.

The value of vagueness can be grasped by a case study that Crenshaw highlights in her work on intersectionality. In the American case of DeGraffeinreid v. General Motors Assembly Div.,Footnote ¹¹⁰ five black women sued General Motors for their employment policy that perpetuated discrimination against black women. In this case, the district court held in favour of the women’s employers—General Motors. The judge held that while there existed a specific legislative remedy against sex-based discrimination that the women could avail, and a specific legal-remedy against race-based discrimination that they could also avail, there was no provision to combine the two specific remedies. An excerpt from the court’s decision is as follows:

This case shows the problem with defining discrimination solely by specifying characteristics, rather than through using relatively vague expressions that cover the concept of discrimination. As the court was bound by the precise legal language of the legislature, they were not free to grant protection to the black women against the discriminatory policies of General Motors.

The law on intersectional claims has fluctuated significantly since DeGraffeinreid.^{Footnote 112} While some American courts have been receptive to intersectional claims,^{Footnote 113} the problem still persists (albeit in new forms). Even where intersectional claims are allowed, courts have often resorted to what has been called the “sex-plus” approach.^{Footnote 114} This approach sees people as composed of one identity plus another—a Black women is seen thus as a compound of two separate identities, instead of having the single, fused identity of a “Black woman”.^{Footnote 115} When employing this approach, courts end up mechanically adding the precise axes mentioned in a statute in their efforts to account for intersectional discrimination.

This sex-plus approach raises some concerns. First, marginalised groups do not always articulate (or indeed experience) discrimination as a mechanical addition of two distinct identities (here, sex and race) but rather as a single, distinctive experiential reality.^{Footnote 116} These articulations find no place in sex-plus constructions. Second, the sex-plus approach also pushes us to ask how many identities can be added together in these hypothetical equations that courts are producing. Courts have briefly considered this question—leading them to restrict the number of axes that can be added together in intersectional claims.^{Footnote 117} Therefore, the sex-plus construction is an inadequate interpretive tool. The complexities of intersectional discrimination cannot be captured through fixed mathematical constructions. As Catherine MacKinnon tells us, the intersectional method cannot truly be realised through “universal generalizations, classifications or abstractions in the clouds”.Footnote ¹¹⁸ An intersectional method must be grounded in the lived experience of hierarchy,Footnote ¹¹⁹ above all else.

How can legal systems, then, account for lived experiences in dealing with intersectional claims? It is important to recognise that the flawed sex-plus interpretive approach is, in part, a result of the linguistic priority accorded to precision over vagueness at the stage of law-making. When statutes list precise axes of discrimination, courts end up perfunctorily adding these axes together—leading to flawed interpretive approaches like the sex-plus construction. The use of precise statutory language proves to be a significant interpretive hurdle in the adjudication of intersectional claims. Therefore, while feminist reform has largely been (and should be) targeted at changing the interpretive approach taken by courts, the use of precise language at the stage of drafting anti-discrimination laws also merits serious questioning.

If we ask these questions, vagueness becomes increasingly relevant for discrimination law. Discrimination law theorists are confronted with the daunting task of defining discrimination—and in doing so, accounting for intersectionality.^{Footnote 120} This debate has arisen, for instance, in the context of Article 15 of the Indian Constitution. Article 15 lists characteristics based on which discrimination can occur.^{Footnote 121} By laying down a list of protected characteristics, the Constitution makers require all newer forms and types of discrimination to fit within or resemble one of the specified axes.

This precision comes at a cost. Indian courts, for instance, have been reluctant to grant that unequal treatment based on sexual orientation can be labelled discriminatory, as such treatment does not fit neatly within the ambit of “sex-based” discrimination that was precisely envisaged under Article 15.^{Footnote 122} Further, Article 15 has been unsuccessful in dealing with claims of intersectional discrimination.^{Footnote 123} As Shreya Atrey argues, intersectionality has had to “manoeuvre the roadblock of the constricted view of discrimination as based only on a single ground” incorporated within the precise wording of Article 15.^{Footnote 124}

Therefore, precision often has a damaging effect in laws relating to discrimination. Vagueness serves an important function in this domain. Tarunabh Khaitan, takes up the task of defining discrimination in the Anti-Discrimination and Equality Bill 2017, that was tabled before the Indian Parliament.^{Footnote 125} Khaitan’s definition incorporates vague predicates, as a necessary part of any definition of discrimination.

Khaitan defines the kind of characteristics that can be used as the basis for discriminatory action. These are “protected characteristics”. Section 3 of the Bill first provides a list of protected characteristics and then, proceeds to include:

This definition is interesting as it does not merely list out axes for discrimination but goes further to allow for judicial discretion in cases by accommodating vague expressions such as “effective control”, “fundamental choice” and “substantial disadvantage” in clause (ii).^{Footnote 127} The use of vague expressions like these could be criticised for being too broad and flexible. However, it is only through a definition that is, to some extent vague, that intersectionality and social discrimination can be adequately dealt with.

Precise laws that seek to neatly define the axes on which discrimination occurs end up creating arbitrary and unjustifiable results as they cannot account for the complexity inherent in certain areas of human life. While in the context of care ethics, we dealt with the complexity that arises out of subjectivity, in intersectional issues we deal with the complexity that comes out of ‘location’.^{Footnote 128} Location, like subjectivity, has to be dealt with through vague predicates and expressions. Precision cannot effectively capture the layers and complexities of discrimination.