Constructing an Argument

Philosophers use a number of tools in constructing arguments. As a philosophy student, it is your task to identify these tools at work in philosophical arguments and analyse how well they have been used, as well as using them yourself.

Amongst the tools with which you will need to become familiar are definitions, distinctions and conceptual analyses. We will show you how to identify each, and the appropriate ways to criticise them. Learning how to use these tools properly will both teach you how to use them in your own writing, and how to spot problems in others’ arguments.



A definition explains what a term means. In philosophy, we call the term being defined the definiendum and the phrase which defines that term the definiens. For instance, if I define a bachelor as an unmarried man, “bachelor” is the definiendum, and “unmarried man” is the definiens.

Philosophers sometimes use this kind of suffix in other contexts as well—for instance explanandum (a thing being explained) and explanans (the thing doing the explaining).

There are two main kinds of definition, which philosophers call intensional and extensional. An intensional definition explains what a word means. For example, I might define “bank holiday” as “a day in the UK on which banks close, and most workers are given the day off”. An extensional definition gives the full list of things which make up the definiendum—for instance, to give an extensional definition of “bank holiday”, I would give the full list of UK bank holidays—“Christmas, Easter Sunday, Good Friday, New Year’s Day, May Day … etc.”


When to define a term:

You should define a term if:

  1. It is a technical term which a reasonably educated reader would not be expected to already know.

e.g. “Definiendum”

  1. It is a technical term being used in a specific sense.

e.g. “realism” being defined as ‘the view that scientific theories are at least approximately true’.

  1. It is a term which has several different meanings, being used in one of those ways.

e.g. “universe” being used in the sense of ‘everything which can be observed’ as opposed to ‘everything that exists’.

  1. It is a new term which the author has introduced. (sometimes called stipulation)

e.g. “I stipulate that ‘appleist’ means ‘discriminating against people who use macbooks’.”

Avoid giving lots of definitions at the start of your paper. You may need to define some things as you go along.
Back to top


Conceptual Analysis

Conceptual analysis is related to definition. Philosophers are often interested in the meaning of concepts—for instance ‘science’, ‘death’, ‘truth’, ‘reality’, ‘art’, etc. To expose the meaning of a concept they sometimes perform conceptual analysis. This technique requires us to provide the necessary and sufficient conditions for X to satisfy a particular concept.

For instance, suppose we want to know what ‘death’ means. We might try to provide a conceptual analysis of death. Usually, instead of trying to analyse the noun itself, we try to analyse the adjectival form (e.g. instead of trying to what “reality” is, we analyse what makes something “real”; instead of analysing the term “death”, we try to state what makes something ‘dead’).

X is dead if and only if X is no longer capable of consciousness and X’s body no longer functions.

When you see the words “if and only if” (sometimes abbreviated ‘iff’), the author is trying to provide necessary and sufficient conditions.


What are ‘necessary and sufficient conditions’?

A is necessary for B when B can only occur when A occurs. In other words, a necessary condition for B is a requirement which must be fulfilled in order for B to take place.

For example:

“In order to enter the library, you must have a library card.”

In this case, having a library card is a necessary condition for entering the library. Necessary conditions can be phrased as “only if” sentences: “You can enter the library only if you have a library card.”

But, crucially, having a library card is not a sufficient condition for entering the library. Just having a library card isn’t enough for you to be able to get in. (It is also necessary for the library to be open, for example).

So, necessary conditions state what must be the case in order for B to be able to happen, but don’t guarantee that B will actually happen.

A is sufficient for B when B must occur whenever A occurs. Sufficient conditions guarantee that B will occur. If we know that a sufficient condition for B has happened, then we know that B has happened.

For example:

“If you have a billion pounds, then you are rich.”

Here, having a billion pounds is sufficient for being rich. If you have that much cash, then you must be rich. Necessary conditions are often phrased as “if” sentences: “You are rich if you have a billion pounds”.

But again having a billion pounds isn’t necessary for being rich. You could be rich without having that much money, or you could be rich in assets. Everyone with a billion pounds is rich, but not everyone who is rich has a billion pounds.

A is necessary and sufficient for B when B always and only happens when A happens. A necessary and sufficient conditions combines the properties of necessary and sufficient conditions. So they are often called if and only if (or ‘iff’) statements.

The conceptual analysis of ‘death’ we attempted above was an attempt to provide necessary and sufficient conditions:

X is dead if and only if X is no longer capable of consciousness and X’s body no longer functions.

If our analysis is correct, then everybody who meets these two criterion (no consciousness and no bodily function) is dead, and nobody who does not meet both conditions is dead.

Definitions take the form of ‘if and only if’ statements.

For example, the definition:

A bachelor is an unmarried man.

could be re-phrased as: “X is a bachelor if and only if X is an unmarried man”.



Necessary = “only if”

Sufficient = “if”

Necessary and sufficient = “if and only if”


How to evaluate conceptual analysis:

When you read an attempted conceptual analysis, like our analysis of death, you should ask two questions:

  1. Is each part of the analysis individually necessary?
  2. Are the parts of the analysis together sufficient?

There were two parts to our conceptual analysis of death—permanent loss of consciousness, and loss of bodily function. Take each question in turn:

  1. Is permanent loss of consciousness necessary for death?
    Is loss of bodily function necessary for death?

If you can think of any cases in which an individual was dead, but had not permanently lost consciousness, or was dead but had not lost bodily function, then the analysis contains a condition which is unnecessary. When this happens, we say that the analysis was too strong. It excludes some cases which it should include.

In this case, one might argue that loss of bodily function is not necessary for death, as some people who are dead have had parts of their body kept functional by means of life-support machinery—for instance, the case of Robyn Benson, who was brain-dead, but whose vital systems were kept functional so that her fetus could be allowed to develop in utero.


  1. Is permanent loss of consciousness and loss of bodily function sufficient for death?

If we can think of any cases in which a person has lost consciousness and bodily function but is not dead, then the conditions are insufficient. When this happens, we say that the analysis was too weak: it includes some cases which it should exclude.


It is possible for a conceptual analysis to be both too strong and too weak—some of the criteria are unnecessary, and the criteria together are insufficient. Some things are included in the concept which should be excluded, and some are excluded which should be included.

To criticise a conceptual analysis, always think:

Are there any cases which the analysis excludes, but shouldn’t? (Too Strong)

Are there any cases which the analysis includes, but shouldn’t? (Too Weak)

The next step in a good critique of a conceptual analysis is to try to develop the analysis. Try to fix the problems you have identifying by changing or tweaking the criteria. Can you fix the problems you have found?
Back to top




Philosophers often make distinctions. They divide a set of things into two or more categories. For instance, some ethicists try to distinguish “moral actions” from “immoral actions”.

Drawing a distinction consists of providing some criteria by which we can sort our subject matter into categories. For example, suppose we want to distinguish between different kinds of animal. We provide a number of categories (e.g. mammal, bird, reptile, etc.) and place our subject matter (animals) into these categories according to some criteria (e.g. “Mammals are warm-blooded and produce milk for their young”; “Reptiles are cold-blooded and scales, not fur or feathers”, etc.)

The goal when making a distinction is to provide a proper distinction.
A proper distinction is both:

  1. Exhaustive
  2. Mutually exclusive

“Exhaustive” means that everything in the subject matter falls into at least one category. Nothing is left out.

Suppose we tried to distinguish between morally obligatory and morally prohibited actions. This distinction would not be exhaustive, as there are some actions which are neither obligatory nor prohibited—they are acceptable but not obligatory.

“Mutually exclusive” means that nothing falls into more than one category.

Suppose we distinguished between “actions with good consequences” and “actions with bad consequences”. This distinction would not be mutually exclusive, because some actions have both good and bad consequences, so would fall into both categories.

A proper distinction is both exhaustive and mutually exclusive—so everything is placed into one and only one category.


How to evaluate a distinction:

If you see a distinction being drawn in a philosophical paper, you should ask two questions:

  1. Is it exhaustive?
  2. Is it mutually exclusive?

Look for things which fall into multiple categories, and things which do not fall into any category.

If a distinction is not exhaustive or mutually exclusive, then a philosophical argument based upon it might fail. A common form of argument based upon distinctions is a dilemma (see below). Dilemmas based on distinctions require that the distinction be at least exhaustive, and sometimes that they be mutually exclusive too.
Back to top