2.6.1 Ensuring coherence in the constitutional provisions

The objective of a constitution-making process must be a constitution that is acceptable, workable, and just. “Workability” of a constitution includes whether the design will match the circumstances of the country, as well as questions of the capacity of the country to operate it, its likely cost, whether it will generate a great deal of litigation, and so on. Less important perhaps are other concerns: readability, length, and durability.

“Coherence” (the various aspects all fitting well together) is a bit different because it covers a number of points, some of which may affect workability, while others may affect acceptability, or even be a matter of style.


A constitution may be “incoherent” in various ways, including the following:

  • The “working parts” (about machinery and procedures) may not reflect the philosophy set out in parts such as the preamble and discussions of national values and human rights.
  • Different parts of the machinery of government may not be suited to work together, particularly where the objective is to set up checks and balances between different parts.
  • It may not even be possible for certain parts of the machinery to work together—this may be a question of dates not matching, or some other fundamental incompatibility.
  • Drafting styles may differ markedly between different parts.
  • There may be great detail in some parts of the constitution and guidelines at the level of principle in others.
  • Words may be used differently in different provisions, or different words may be used for the same concept or institution. This is not just a matter of style; it may affect how the constitution is understood and applied.
  • There may be repetition—which may cause confusion, difficulty in reading, and also problems of comprehension.
  • The document may be so badly organized that it is hard to find specific points, and the reader runs the risk of failing to realize that what she is reading is restricted or expanded by a provision somewhere else in the document.

Box 31. Parliament as the source of incoherence in the Fiji constitution

In Fiji in 1997, a parliamentary committee took over the final stages of the constitution- making process, after receipt of a full and coherent set of proposals by a commission, and introduced a power-sharing arrangement (which had already been rejected by the commission) under which any party with a significant number of parliamentary seats could take part in the cabinet. The committee members failed to think through the implications for other aspects of the constitution. (And time pressures made it hard for anyone, including professional advisers, to realize the problems that had been created.) A notable example involves the provisions on the senate. To ensure that body included a “nongovernment” element, the original proposal was that the leader of the opposition would nominate some senate members. This was retained by the parliamentary committee. But at the same time the constitution said that those senators must be from parties entitled to sit in the cabinet. So they were to be nominated by the leader of the opposition, but not to be members of the opposition, and not to offer a nongovernment voice.


Incoherence can arise from various causes, including:

  • civil society focusing on “values” while politicians focus on power, and experts on “workability”;Part 2: Tasks in a constitution-making process
  • making changes at late stages in the process;
  • a tendency, on the part of those preparing the constitution, to focus on only one part of the document at a time, or even on only one article at a time;
  • dividing the preparation of proposals among committees in a way that leads to overlap;
  • asking nonexperts to produce actual drafting language for the constitution, rather than the ideas;
  • uneven knowledge on the part of those involved: it is almost inevitable that a body composed of those with some, but not great, knowledge of constitutions (and few countries can command the services of many experts with great knowledge) may seize on some idea without being able to understand all its implications, and even without understanding how it has worked elsewhere;
  • cutting and pasting from a variety of foreign models;
  • having different drafters with different styles working on different aspects;
  • excessive speed; and
  • a reluctance to use experts, and a certain arrogance on the part of those (usually politicians but sometimes lawyers) who think they have the skills and the mandate required.


Avoiding incoherence

Careful planning of the process—which is essentially the theme of this entire handbook—should avoid many of these risks. The planning needs to be done not only by those officially in charge of the process, but also by those wishing to influence it, including political parties and civil society generally.

Specifically, the following strategies can help avoid the various pitfalls outlined here:

  • having in advance, or adopting at an early stage, a set of guiding principles (see part 2.1.8); however, these are likely to be broad, while issues of coherence are more likely to involve detail;
  • civil society focusing not only on the values and rights, but working hard to see how the difficult technical aspects can achieve their aims;
  • accepting the necessity for skilled and technical expertise at the decision-making stages, even for carrying into effect the necessary political compromises;
  • having a skilled “harmonization committee” (usually found within a constituent assembly) whose responsibility is to put together the various elements of the draft and ensure coherence in the whole;
  • minimizing the number of technical drafters, and ensuring that they are both competent and familiar with the style to be adopted; having one person in charge of the drafting with the authority to instruct drafters to use certain phrases and styles; and developing a manual of style for the particular process;
  • preparing an explanatory glossary of terms to be used, including translation into major languages, and ensuring that it is used;
  • having a workable timetable;
  • having a philosophy of work that avoids procrastination, assuming that the timetable is to be taken seriously. Otherwise, important stages toward the end are squeezed; this is an important role, especially for the chairs of the bodies in charge; and
  • avoiding superficial knowledge by organizing in-depth study groups on particular issues, including inviting experts, local and foreign, and ensuring that study tours do involve a rounded experience of the topics studied.

2.6.2 Drafting the constitutional text

A constitution is a political and a legal document. Ideally it should be:

  • understood by the people;
  • usable by politicians and bureaucrats; and
  • able to be interpreted by the courts.

All readers should find the same meaning in the document. The aim of the drafter should be predictability; there is none if the people, the politicians, and the courts come to different conclusions about its meaning.

This can be qualified to a limited extent—the people may take away from their reading of the constitution something different than do the lawyers. But those additional meanings should not be about the way the machinery of government operates. It will instead involve a sense of whether this document is “their” document—does it speak to them? Does it have any resonance for them?The art of the constitutional drafter is to connect with all these readerships. But any lawyer would insist that resonance with the people must not stand in the way of maximum predictability. Unlike many laws, a constitution has a variety of different components, some of which lend themselves more than others to reaching out to the people.

Occasionally it is intentionally left to the courts to decide what a provision means (as in the example from South Africa mentioned above [see part 2.2.3], where although the constitutional assembly could not agree on a position on the death penalty, the issue was left to the courts to decide whether the “right to life” affected the death penalty).

The courts’ approach

Diana Yankova has written that the “draftsman always has to keep in mind that he is writing for a hostile audience—his text will be interpreted by warring sides in the courtroom” (Yankova 2006). The people may find it hard to understand the need to express an idea in a way that the courts can understand, but lawyers will want to use precise language that sometimes seems rather dry, and to use the words used in other constitutions. Managers of a constitution-making process must recognize the importance of careful use of language—even in a process that is people driven. To take a simple example: in the Kenyan process, environmental enthusiasts seized on the notion that at a minimum Kenya should have 10 percent tree cover (a figure below which it had fallen considerably), and drafted a provision that “every piece of land must have 10 percent tree cover”—this in a country with semidesert areas.

When interpreting a legal document, the courts make certain assumptions, such as that:

  • language used before is being used in the same sense; and
  • a change of language is intended to convey a change in meaning.

They look at how courts elsewhere have interpreted similar language, and at international law if that is where the language comes from. And they will be concerned about detail:

  • “includes” is not the same as “means”; and
  • “shall” or “must” is not the same as “may.”

There are often well-established traditions about how certain aspects of a constitution are used by the courts. In particular, the preamble and any “directive principles” will often not be treated as rules—though they may be used as guides. Citizens may find these parts of the constitution easiest to read, but might be disappointed to discover that in the courts these parts have limited legal force.

Drafting styles

The tradition of legal drafting in continental Europe, and in countries that have adopted European traditions, is broader, and interpretation focuses more on the spirit and less on the actual words of a legal text, while the common law tradition (used in England and countries that were once British colonies) is more detailed. Constitutions in the United States tradition also tend to be more concise. But differences are much less marked in modern constitution- making because:

  • of participatory constitution-making;
  • of the influence of international law;
  • of occupying forces, international agencies, and international NGOs;
  • of constitutional borrowing; and
  • even in common-law countries, constitution drafting is less rigid and complex than ordinary statute drafting (because a constitution is as much a political as a legal document).

In some countries (or for some judges) it is common to look at the discussions that took place when the law was being made; in some countries this is unusual, or even limited by the law.

Stages of text development

There are three “intellectual stages” in producing a constitutional provision:

  • adopting the idea (a job for the people with political, legal, and other expertise);
  • framing the provision in a legal way (a job for constitutional lawyers, with political scientists and other technicians); and
  • preparing a final legal text (a job for legal drafters in close collaboration with a drafting committee).

This neat scheme of things—used in many national lawmaking processes—may not apply in a constitution-making process. A draft of a final text is often produced at an early stage, or else those involved resort to existing texts rather than beginning with the ideas, because:

  • lawyers or active politicians find it easier to operate with familiar existing provisions;
  • a political party or group produces a draft;
  • advisers (local or foreign, national or international) offer concrete suggestions;
  • a committee or commission working in advance to generate ideas for the constituent assembly does so in the form of a draft constitution; or
  • time pressures exist.

And the technical drafters may be involved at a much earlier stage than is usual in national lawmaking. This is a good thing—so that drafters have an accurate idea of what the people want from the constitution, though it may call on the drafters’ reserves of patience as they have to interact with the people and their demands instead of only carefully formulated drafting instructions. The drafters may need to understand the culture and the symbols of a particular society.

Box 32. Development of the text in some modern constitutions

Fiji: The Reeves Commission, which produced the proposals, included no lawyers, but had two legal advisers. The report includes suggested language for many provisions. But the final drafting was done by a drafter in Australia.

India: The 1935 Constitution of India Act was an important basis for the work of the constituent assembly. The constitutional adviser, B. N. Rau, a historian, judge, and capable drafter, prepared a draft on the basis of the reports of major committees, making use of the 1935 act, the U.S. Constitution, the Irish constitution, and others. A drafting committee (which Rau attended) worked on this further; the drafting committee made the final changes after debate in the full assembly.

Kenya: Legal drafters (from several countries) were involved in the preparation of the first draft, working parallel to the commission. Some provisions were derived from the Ugandan and South African constitutions. Drafters assisted most committees, and then revised the draft in the light of the constituent assembly’s decisions.

South Africa: The interim constitution of 1993 was negotiated largely by parties; the 1996 constitution was prepared through the constituent assembly. Both processes used committees of technical experts. A number of drafters worked on the final constitution, and one was appointed with the mandate to produce a plain-language, gender-neutral document.

Uganda: The commission set up a drafting committee, and also used the services of two professional drafters.

Vanuatu: There was no legal framework for the constituent assembly, and most of the drafting was done by the legal adviser to the local political parties (a nonlocal), with contributions from an experienced Papua New Guinea lawyer.

The work of the technical drafter

Legislative drafting is a specialization within the field of law. Most people who become drafters have experience with legal practice, and they will have special training in drafting. Their usual task is to convey what their clients want as clearly as possible, and in “judge-proof” words (that is, words that the courts will understand in the same way the drafters do). In some legal traditions, notably those of Europe, there is no profession of legal drafting; drafting is usually done by legally qualified civil servants.

In constitution-making, drafters have a more demanding task. Understanding what the clients want may be harder when those clients are the people rather than a ministry’s lawyers. A constitution also touches on all sorts of other laws, and the drafters will need to be aware of these implications. Finally, it is particularly important that a constitution be written in clear, even eloquent language. It is important to understand that a drafter is not concerned only with the words. One has said, “To express an idea clearly within the constitution the drafter must first understand the idea and its place in the universe of constitutional ideas.” This sort of drafting is the art of:

  • applying imagination, logic, and reason to legal problems;
  • matching policy choices to functional purposes;
  • designing legal instruments to achieve those functions;
  • crafting each legal sentence in a formulation appropriate to its particular function; and
  • choosing the best words to communicate those functions effectively and economically.

Drafting the constitutional text

It is tempting for people involved in preparing a new constitution to look at other countries and “cut and paste,” and a lot of this occurs. But it is important to understand not only that a constitution must be understood in the particular historical and political context of the country, but that there are different styles of drafting. We might distinguish:

  • the “what is technically necessary” approach; this is the traditional approach, of including only legally binding provisions, and drawing a clear line between what is appropriate for a constitution and what should be reserved for “regular” legislation;
  • the “didactic” approach, which is how one might describe the approach used in Papua New Guinea, where the drafters were aware that running a government was something that most Papua New Guineans had no idea about; the result is a constitution that may be the longest in the world (for a few million people) but is a detailed “user’s manual” for government;
  • the “explanatory principles” approach, which was used in South Africa, where many of the chapters are prefaced by some principles designed to show the politicians and the officials, and the public, the purposes of the provisions;
  • the “reaching out” approach, in which the language is directed as much to the people as to the institutions and individuals charged with operating the constitution; and
  • the “let’s put it all in” approach, which has been used in some Latin American countries; it offers great detail, especially on some current issues such as the rights of indigenous peoples (and might be used to describe the constitution probably forthcoming in Nepal).

Differences may arise due to local political and legal traditions, to citizen involvement in the process, to the use of international instruments, to degrees of trust or mistrust in institutions, and to individual drafters’ input. Although there is a general trend toward similarity across countries, traditions, expectations, and practices may change in any given country.

Issues in drafting

The following points will not be new to drafters, but they should be remembered by those who have to explain what they want to drafters, and have to try to understand the drafters’ art.


Most countries have a small number of official languages—often only one—and laws may not even be translated into all of these languages. The constitution, as the basic law of a country, is one that ought to be in as many languages as possible. Drafters find that the need to be able to translate affects—and simplifies—the style of drafting in the original language.

In order to help drafters, it will be necessary to develop a glossary of words so that the same word or phrase is always translated in the same way.

The need for legal certainty means that one version of a law must be the authoritative one. Often this will be the one in force when the drafting was done, because it is then that the expert legal input occurs.

A decision will have to be made about when to translate, and whether to translate all drafts. It may be wise to translate the existing constitution for use as a civic education tool. And if this is well done, it will help in future translations. (See part 2.3.9.)

“Plain English”—or other languages

Modern trends have been toward using simpler language in legislation. In South Africa [1996] a legal drafter was appointed specifically to ensure that the final draft was in “approachable” language. But that drafter insisted that although “simplicity, precision, and clarity” are desirable, they are only tools for the achievement of the overarching objective of predictability.

Inclusiveness and approachability

Especially in the English tradition of drafting, there was a “masculine” tendency in constitutional language, and the word “he” was to be read as including “she.” It is possible to draft in ways that avoid this clumsy and sexist tradition. Other languages do not raise the same issue, but there may be other ways in which language can convey certain assumptions about who holds office. The word “president” in Hindi and Nepali literally means “husband of the nation,” which has been controversial. Language carries all sorts of cultural implications, and the drafters need to be sensitive to them.


Many people argue that short constitutions are better. They should not be longer than needed; excessive length may discourage careful reading, and unnecessary words may interfere with interpretation. But length alone should not be the test; each provision should be there for a reason. Elements that make some constitutions longer include:

  • more topics covered;
  • more ideas;
  • more words than needed;
  • the role of the constitution as the “manual for government” approach—as in Papua New Guinea (see appendix A.9) and some recent African constitutions; and
  • in the case of a federal system, the absence of separate state constitutions—contrast India, where the state constitutions are included in the national one, with the older federations such as Australia and the United States, where state constitutions are separate.

Constitutional structure

There are ways to organize a constitution that make it easier to understand, including putting details (especially if they may be elaborated in an ordinary law) into schedules (appendices). In some countries (e.g., Papua New Guinea) the constitution provides for special laws to address the details of government, rather than putting everything into the constitution itself. These can include elections and courts. They can be called “organic laws” and be harder to change than ordinary laws, though not so hard to change as the constitution itself.

The very order of a constitution’s chapters may be significant: is the president more important than the parliament? Are rights less important than government?

Risks and dangers

Things may go wrong at any stage in a process as complex as preparing a constitution. Even the technical drafting aspect has its risks—for the process and for the drafters. Spoilers or self- interested individuals may try to influence the drafters, to delay the process or to change the text. Precautions may be taken to:

  • make sure that it is perfectly clear what has been decided in terms of substance;
  • make sure that it is clear from whom the technical drafters are to take instructions; and
  • guard against the risk that the “master” version of the text might get lost or be interfered with.

Some issues for the drafters themselves

This book is not written specifically for drafters, who will have their own professional expertise and experience to draw on and will go to more detailed resource materials. However, here are some important points, drawn from the writing of experienced drafters:

  • It is not the role of the drafter to be the “constitutional visionary, negotiator or advocate” (Knight 2008).
  • The drafter’s client is not just the group from whom instructions come, but the nation.
  • A constitution should be coherent as a document, and last-minute changes should be avoided, as should cutting and pasting from other constitutions without regard to the unity of the whole (Lynch 1988).
  • Constitutions should say what they mean, and not rely on conventions and fictions, especially in countries not used to them (for example, one should not say that the head of state must act on “advice” when the intention is that the advice must be followed) (Lynch 1988).