Is there more to freedom than absence of external constraints?

In this essay, I will be argue that there is more to freedom/liberty (interchangeable throughout) than
absence of external constraints. I will use an analogy of sentient robots to explore the idea of various
constraints on liberty. I will link this to a discussion of mixed cases, such as the oppression of women,
where external and internal constraints cannot be easily separated. Then, I will try to isolate internal
constraints and argue for why they, too, can constitute a violation of liberty. Then, I will assess the
normative significance of liberty, and attempt to show – paradoxically – that liberty can only exist in
presence of constraints.
Imagine a sentient robot with access to unlimited resources. Its desires had been programmed
beforehand, but otherwise it is perfectly human. It has no external constraints on its desire satisfaction,
as it is far too powerful to be constrained by anything on earth, be it agent or accidental, from fulfilling
the desires that it has programmed. Is it free? It had no agency in what its desires were, though it had
agency to act or not act upon them. Its behaviour is also not constrained by obstacles other than that
programming, thus, necessarily, is absent of external constraints. Yet, it has no way to affect its
actions other than acting contrary to its desires. Because of this, the robot clearly isn’t free, or at least,
not as free as it could be.
If the programming of the robot was to be denied status as an external constraint, it would be easy to
answer the question there and then. The programming is external and a constraint. This situation
changes however, if the robot could overwrite its programming, but chooses not to, as it is inconsistent
with its present desires. While the constraint exists, the robot chooses to follow it, rather than use its
agency to act against its present desires to craft new ones. So the constraint doesn’t really constrain the
robot – it is bound by its own will as much as by its creator. While this example might seem
convoluted, a feminist reading of freedom, offered by Hirschmann, can offer insights into how this
mechanism is reflected in the real world.
The real world is structured along the lines of the patriarchal past, where oppression of women was
sacrosanct and legally enshrined. In the days since, most of legal patriarchy had been dismantled. The
external constraints, such as sexist laws, no longer oppress women to the degree that they did just two
centuries ago. But, even if every sexist law was repealed, the issue persists. The prevailing structure of
legal patriarchy left a scar on the equality of the men and the women – as can be seen in police
frequently ignoring cases of sexual assault. And while this discrimination can still be seen as purely
external constraints, what of the case of a battered wife who returns to her abuser despite no evidence
of him changing? In childhood, values of submission and obedience were taught to this woman as
desires. Now, though it would be possible to override her desires, she chooses not to – as remaining a
“good wife” is preferable to her. The case is analogous to the robot case.
In other words, when we ask the question, “is there more to freedom than absence of internal
constraints?”, the question can be rephrased as “in a world where all external constraints were taken
away, would this vacuum be filled with perfect freedom?”. The answer to the second question is no –
external constraints manifest themselves as internal constraints too, preventing people to act in ways
other than ways to which they have been conditioned. If we accept that the robot in my experiment
isn’t free, then we ought to accept also that those whose internal constraints come from external
constraints are not.
Let’s now consider an agent who has a set of internal constraints which are not caused by any kind of
oppression, or external deprivations. Let’s say, someone who had freely chosen to take drugs out of
intellectual curiosity, and liked the initial experience so much they continued until they couldn’t stop –
developing an addiction. This is not an addiction caused by them being externally constrained to act
the way they do – no peer pressure, work stress, or anything that would make this case similar to the
cases of oppression-generated internal constraints. Their desires have since shifted, as they don’t enjoy
the use of their drug anymore. Yet, they cannot stop – their addiction stops them from ending it.
Berlin tries to separate liberties into two types, positive and negative. Had the current drug addict been
prevented, perhaps legally, from taking drugs the first time they did, would violate their negative
liberty. Their current condition, where the real will is prevented from acting due to the drug addiction,
violates their positive liberty. One reading of Berlin’s account can be interpreted as a statement of
preference – it is better to violate the positive liberty than to violate the negative liberty, as positive
liberty considerations are used to justify tyrannical governments and despotic interference of the state.
The normative preference doesn’t follow – all that Berlin shows is that negative and positive liberties
should be seen as different from one another. However, the implicit argument is that positive liberty is
not a liberty at all, and in fact it is just a component for desire satisfaction, where the desires satisfied
are other than liberty. However, Berlin’s account falls apart when one considers that both positive and
negative liberties lack normative significance when considered as ends in themselves, and should both
be read as tools for achieving other ultimate ends. I will develop this case in a later part of the essay.
For now it suffices to say Berlin’s account does not provide us with an answer as to what would be
preferred in this case from a liberty maximising point of view. But Taylor’s account might.
Taylor shows how various goals, for example a trip in the Andes, are important for normative reasons,
as self-actualization is such a normatively important reason. If a person, such as our addict, was to be
prevented from such self-realization, one would be less free than someone who isn’t. A major internal
constraint such as a drug addiction could take away the agency to self-constitute, and therefore make
this person unfree to pursue one’s goals. Here, freedom is understood as possibility to pursue one’s
goals and ambitions without obstacle. Taylor discusses how an external constraint of stopping at a red
light, violation of negative liberty, is not a violation significant enough to constitute a normative harm.
So in this case, we have a minor loss of negative liberty and a major loss of positive liberty – similar
case can be done about the drugs from earlier. We impose a minor violation of negative liberty by
taking away the right to freely consume drugs. But in effect, we prevented a much greater violation of
positive liberty, which would arise from the drug addict being unable to act against their addiction.
Under Berlin’s account, this would be a net loss of liberty, and a trade-off into another value, like
security or happiness. In reality, a man whose only constraint is a ban on using drugs is freer than a
man who cannot stop taking drugs, regardless of material status or desires. This much is self-evident
under any conception of freedom, even if it’s understood only as freedom from constraints.
And since in this case, maximizing liberty leads us to impose external constraints, which when taken
effect will increase, not reduce liberty, the case for freedom as absence of external constraints and
nothing else falls apart.
Before I reach the boldest claim of my essay – that liberty is not normatively important – I want to
briefly discuss how the distinction between negative and positive liberty can itself be fallacious. This
relies on a simple linguistic trick exposed by MacCallum – constraint of freedom can be transformed
into lack of a freedom to without much effort. An example is a chained prisoner – is he constrained by
his chains? Or is it that he lacks freedom to use a key on his chains? Both interpretations are
reasonable. Thus, I will collapse positive and negative liberty into liberty for the next part of the essay
for simplicity.
Let me first explain how liberty lacks normative value on its own.
An artist uses her freedom to advance herself and humanity at large. The soulless billionaire uses their
freedom to gain more money. Does the fact that they are both free carry equal normative meaning?
How does it change if the action of gaining money harms others, though not by depriving them of
freedom, but by causing temporary physical pain? Is it good that the soulless billionaire is free? If
liberty was a normative value on its own, regardless of consequences, we would have to say yes. I’d
argue it’s the opposite. The freedom of the billionaire causes pain. The pain doesn’t violate liberty – as
no one is constrained by this pain, since it’s temporary – yet it would be better for the billionaire to be
unfree to pursue that goal. The goal is wrong, or at least it isn’t normatively good. This is what makes
liberty wrong here, not a conflict of liberties. And since there is a case where a net reduction of liberty
is good, liberty cannot be a positive normative category. On the other hand, if the artists liberty was
constrained and prevented her from creating art, this would be negative – self realization has a
normative value, and liberty is a tool to such realization. In her case, taking away liberty would be
normatively wrong. So, liberty has both negative and positive normative value. And so, it doesn’t have
normative value at all – otherwise, it would be contradictory.
Yet, in both those cases, presence and absence of liberty was normatively important. How can it be if
liberty isn’t normatively important?
This stems from liberty being a key component in achieving goals that are normatively valuable. There
are various normatively important goals. One such goal, as shown by Taylor, is self-realization. There
are others. Mill’s utility serves as the most intuitively accepted of those goals. It is preferable to have
utility rather than not, and while it is preferable to have the utility maximising outcome arise without
compelling people into anything, this is because of self-realization, not because of importance of
liberty itself. Mill admits in his writing, that those who are unable to reap the fruits of liberty will not
benefit from it, such as underage people, or people who are developed to such a low level that they
may well be children. Those people would not benefit from liberty. Fully developed people can either
take positive normative value or negative normative value from their liberty. The value of liberty is
coloured by its end goals and end achievements. Berlin’s account isn’t convincing enough to establish
a counterpoint to this claim – and all the other accounts I mentioned have to be understood as
advancing this idea, if not explicitly, then implicitly. 1
Finally, due to time constraints, I will limit my exploration of how liberty only exists in presence of
constraints to a short paragraph. Suppose we return to our sentient robot from earlier, but now suppose
it did reprogram its desires so that it desired nothing at all, to free itself from constraints of desire. Yet,
is it free? There is no way to assess its freedom, as there are no constraints. The actions that it takes
are unconstrained by anything, including desires. So it can act in whatever random, unpredictable way
it chooses. Its freedom can never be assessed, as it never comes into contact with constraints.
Intuitively, my reading of the situation suggests the robot is not free, or at least, it’s freedom is not the
same kind of freedom it could have had under constraints. Thus, there is a significant connection
between the action of constraining action and liberty, though exploring that would need an entirely
separate work.
Summarising – liberty is not just freedom from external constraints. There are internal constraints
which also deprive an agent of their liberty. An example is drug addiction. Sometimes, imposing
external constrains is liberty maximising. This means that it cannot be the case that liberty is only a
freedom of those external constraints. The distinction between the constraints itself is quite muddy, as
shown in the feminist account of freedom. And finally, normative value of liberty is null without the
goals that this liberty advances, as a net increase of liberty can both be normatively bad and
normatively good. I was not able to show that liberty is void without constraints, however I wish to
explore this idea more so left it in this work.

1
This doesn’t apply to MacCallum’s account, as I didn’t fully understand what concept of liberty he advocates.