=Paper=
{{Paper
|id=Vol-2518/paper-CAOS1
|storemode=property
|title=An Analysis of the Notion of Need for the
Representation of Public Services
|pdfUrl=https://ceur-ws.org/Vol-2518/paper-CAOS1.pdf
|volume=Vol-2518
|authors=Luca Biccheri,Roberta Ferrario
|dblpUrl=https://dblp.org/rec/conf/jowo/BiccheriF19
}}
==An Analysis of the Notion of Need for the
Representation of Public Services==
https://ceur-ws.org/Vol-2518/paper-CAOS1.pdf
An Analysis of the Notion of Need for the
Representation of Public Services
Luca BICCHERI a and Roberta FERRARIO b
a University of Urbino (Italy)
b ISTC-CNR, Laboratory for Applied Ontology, Trento (Italy)
Abstract. Many Public Administrations structure their services around the no-
tion of users’ need. However, there is a gap between private, subjectively per-
ceived needs (self-attributed) and needs that are attributed by PA to citizens (hetero-
attributed). Because of the gap, citizens’ needs are often only partially satisfied by
PA’s services. This gap is in part due to the fact that the meaning of the word “need”
is ambiguous and full of antinomic nuances. The purpose of this paper is to formu-
late a definition of “need” suitable for citizens’ needs management with respect to
PA’s services offering, and to provide an accurate ontological analysis of the notion
of “need” and the network of concepts that relate to it.
Keywords. need, instrumental need, self-attributed need, hetero-attributed need,
mental attitude, Public Administration, goal, satisfier, service.
1. Introduction and motivation
Public administration (PA) is a constituent element of political action directed towards
the satisfaction of citizens’ needs. The main events of people’s lives such as births,
deaths, weddings, changes of address, are recorded and managed by PA. Every day, citi-
zens contact the PA’s offices in order to obtain identity cards, certificates of professional
competence, services. With regard to the latter, the ongoing procedure of digitising PA’s
data allows for the improvement of efficiency and simplifies the citizens’ access to the
requested service.
However, there is a gap between private needs (self-attributed) and needs that are
attributed by PA to citizens (hetero-attributed). Because of the gap, citizens’ needs are
often only partially satisfied by PA’s services. This gap is in part due to the fact that the
meaning of the word “need” is ambiguous and full of antinomic nuances.
A need can be defined as absolute or relative, objective or subjective, natural or
artificial, endogenous or exogenous, private or social, particular or universal, and so on.
Therefore, there is currently no consensus about what ‘to have a need’ exactly means.
It is thus extremely important to engage in a rigorous conceptual analysis that would
enable to distinguish between various notions of meaning and single out those that are
more functionally useful in analyzing and representing services.
Given the relevance of needs for human well-being, through the centuries different
disciplines, from philosophy to economics, as well as anthropology and psychology, have
been promoting several theories of needs trying to explain their nature. More recently,
the area of information science also got interested in understanding needs and using such
Copyright c 2019 for this paper by its authors. Use permitted under Creative Commons License Attribution
4.0 International (CC BY 4.0).
notion as a guiding conceptual device to represent and manage consumers’ and citizens’
necessities. To this aim, a deep analysis of the notion of need is of paramount importance.
The purpose of this paper is to take some steps forward in this direction, trying to
formulate a general notion of “need” able to provide a focus around which PA’s services
can be structured and made accessible.
2. Ontological analysis of needs and their connections with other entities
The contemporary philosophical debate about needs revolves around two main issues:
the moral one and the ontological one. In short, the former is focused on the problem
of satisfying the so-called “fundamental human needs”. The satisfaction of those needs
should ensure respect for universal human rights and support a more equitable worldwide
distribution of wealth. From this perspective, Manfred Max-Neef, Antonio Elizalde and
Martin Hopenhayn [1, pp. 13-47] suggest that a good economic theory has to answer,
first of all, the following question: what are fundamental human needs?
Actually, first we should be able to know what ‘to have a need’ means, so as to an-
swer the question above. Trying to single out the meaning of ‘having a need’ is precisely
the aim of the ontological issue. Although it is clear that the needs’ moral issue is closely
related to the ontological one, in this article we are going to discuss only the latter. Let
us start to take into consideration the more relevant theoretical elements about the nature
of needs.
In literature, we can find two different ontological perspectives about needs: instru-
mentalism and absolutism. The supporters of the instrumentalist theory believe that all
needs are instrumental, that is, they are necessarily ontologically dependent on goals or
ends or purposes (in other words, needs could not exist without goals). Philosophers who
sustain absolutism, instead, reject this assumption, claiming that although it is true that
some needs are instrumental, others are absolute, namely they can exist independently
from goals or purposes. According to absolutists, there would be an evident conceptual
distinction between instrumental and absolute needs. For example, consider these cases:
(A) Paul needs a camera to take a picture
(B) Maria needs food
(A) shows the typical logical structure of means-end reasoning, where Paul’s need
for a camera plays an instrumental role for Paul’s goal, that is taking a picture. The
general conceptual frame that represents an instrumental need is ‘A needs X for P’, where
A stands for an agent, X stands for a resource (an object or an action) and P stands for a
goal. Both absolutists and instrumentalists agree upon the fact that (A) is instrumental.
Regarding (B), an exponent of absolutism would affirm that the need under consid-
eration doesn’t ontologically depend on other goals. In fact, Maria’s need for food is an
end in itself, as well as other “vital needs”, like water or shelter. McLeod [2, pp. 211-
213] believes that vital needs are universal; for this reason, they are shared with other or-
ganisms (besides, an organism could not have existed without having them). In contrast
to other needs, such as the need to smoke, vital needs are those absolute needs that we
can’t refuse to ascribe to ourselves because they depend on our biological constitution.
Clearly, we can refuse to satisfy them. For instance, Maria can refuse to eat, but if she
doesn’t eat she will be harmed. Wiggins [3, pp. 10, 14] explicitly points out that absolute
needs are those the non-satisfaction of which implies harm.
In any case, instrumentalist philosophers can still explain (B) in terms of the means-
end prospect. For example, it is possible to put forward the idea that the satisfaction of
Maria’s need for food is necessary for avoiding physiological harm caused by hunger.
Therefore, ‘avoiding harm’ would represent Maria’s aim in (B). This interpretation is
based on the elliptical thesis, that is: “A needs X” is a short form to say ‘A needs X for
P’. This is one of the thesis shared by instrumentalists.
However, McLeod [4, 6] warns against the risk of confusing the absolutist theory
with an instrumentalism-by-the-back-door. In fact, he states that:
When an absolutist regards an absolute need as a need that must be met for harm to
be avoided, this does not result in collapse of the absolute/instrumental distinction.
Some beings without ends, goals or purposes of their own, such as plants, can be
harmed.
To sum up, the topic of dispute between absolutists and instrumentalists is based on
the needs’ ontological dependence on goals. The former deny that all needs are necessar-
ily ontologically dependent on goals, ends or purposes. The latter state that ontological
dependency on goals is a necessary condition for something to be a need.
Moreover, various philosophers, in particular Fletcher [5], have attempted to stress
the importance of the notion of “necessity”, so as to shed light on the meaning of “need”.
With regard to the philosophical debate on needs, Fletcher’s position is neither absolutist
nor instrumental. According to him, there would not be a conceptual difference between
instrumental and absolute needs. Rather, Fletcher asserts that there would be different
kinds of necessity in the face of a single concept of need [5, pp. 12-13]1 . Nevertheless,
with the exception of Fletcher, both instrumentalists and absolutists agree upon the fol-
lowing three points:
(i) There are entities called needs
(ii) Needs are different from their satisfiers
(iii) There are different types of needs
Regarding (i), even though philosophers don’t question the existence of needs, they
might disagree about the kind of ontological category that has to be used to classify
them. McLeod [2, pp. 214-220] casts doubt on the so-called “phenomenological thesis”,
denying that ‘a person can know his or her own needs by feeling them’. He suggests that
we should distinguish between a need and the experience of it.
For instance, when someone says ‘I need a drink’, he doesn’t feel the need for a
drink, rather he feels thirsty. So, a need can be indirectly manifested by means of a
1 We believe that the concept of necessity referred to needs should not be intended as the modal logical state
of necessity, that is to say, that something is necessary when it is not possible for it to not be the case. Rather,
the notion of necessity, as Castelfranchi [6, p.57] has stressed, is related to negative perceptions, feelings or
emotions which are linked to needs (e.g., the need for water is associated with the sensation of thirst). So, the
need’s necessity is due to a kind of physical urge which is ascribable to a need’s perception, rather than to
the need itself. Furthermore, usually there are different satisfiers which can satisfy a specific need q in order
to achieve a goal p. This means that the choice of a specific satisfier, far from being mandatory, is optional
(what really matters is that satisfiers allow goal achievement). Note that, from an instrumentalist point of view,
satisfiers are resources and, like all resources, they are limited and require an effort to be acquired. If the
satisfier is not reached, then the related need is not satisfied, and the agent’s expectations about the goal are
frustrated. We suggest that the avoidance of such frustration represents another reason that explains why needs
are especially urgent.
feeling, but the need is not itself a feeling. Needs can exist regardless of feelings. For
example, Paul could have an iron deficiency and not knowing it, because he can’t feel
his iron deficiency. Nonetheless, his physician could inform him about it. So Paul would
be conscious of his need only through a third-person knowledge, that is a physician’s
knowledge. Furthermore, needs, in contrast to desires, can’t be a type of mental states.
Very roughly, if needs were mental states, organisms without mental lives such as plants,
would not have needs, which is unlikely. Therefore, according to McLeod, a need is not
a type of experience or mental state. So, what is a need?
Unfortunately, McLeod doesn’t provide any ontological definition. However, he
strongly states a relevant epistemological issue about the philosophy of needs, which
is related to point (ii). McLeod [2, p. 218] affirms that the source of our knowledge of
needs is neither perception nor introspection, but rather an inferential process that has as
outcome a satisfier. Namely, we can know a need only indirectly, through an inference
about what will satisfy our need (like in the above example of Paul’s need for iron, which
could be detected by medical diagnosis).
It is worth noting that Paul’s need for iron is not iron itself; in other words, we must
distinguish between a “need about x” and “x”, which is the need’s satisfier. A satisfier
can be seen as a resource (object or action) that has the function of satisfying a need.
Generally, there are many satisfiers which can satisfy a specific need, and vice-versa.
With respect to (iii), in literature it is widely recognized that there are many kinds
of needs and ways to categorize them. The most famous hierarchy of needs is repre-
sented by Maslow’s pyramid [7], where needs are ordered according to the principle that,
among needs, some of them would be more fundamental or urgent to be satisfied, com-
pared to others. Maslow’s hierarchy is grounded on five categories of needs, which are:
physiological, security, social, esteem, self-actualization.
Another interesting theory of fundamental needs has been proposed by Max-Neef
[1, pp. 30-31], where needs are classified on the one hand through the existential cate-
gories of being, having, doing and interacting and, on the other hand, according to the
axiological categories of subsistence, protection, affection, understanding, participation,
idleness, creation, identity and freedom. From Max-Neef’s point of view [1, pp. 27-28],
needs don’t change through time, or rather, they ‘change with the pace of evolution, that
is to say, at a very slow rate. Therefore, fundamental human needs are not only universal,
but are also entwined with the evolution of the species’. In contrast to needs, satisfiers
change according to time, cultures, and circumstances.
Finally, in a very influential book [8], Len Doyal and Ian Gough present a distinction
between basic and intermediate needs, in which the former are universal preconditions
that enable participation in any form of life, while the latter connect basic needs to their
fulfillment.
The two basic needs that they identify are physical health and autonomy2 , for which
the “avoidance of serious harm” is implicitly considered as a final purpose. In their
words:
since physical survival and personal autonomy are the preconditions for any individ-
ual action in any culture, they constitute the most basic human needs – those which
2 The inclusion of personal autonomy within basic needs is interesting, as it seems to presuppose a consider-
ation for an “internal” or “mental” dimension that is neglected by other authors who deal with the concept of
basic, fundamental or absolute need.
must be satisfied to some degree before actors can effectively participate in their form
of life to achieve any other valued goals. [8, p. 54]
Intermediate needs are instead conceptualized as satisfiers of basic needs, goals that,
when achieved, contribute positively to the satisfaction of basic needs3 .
The authors also provide a list of intermediate needs, comprising: food and wa-
ter, housing, a non-hazardous work environment, a non-hazardous physical environment,
health care, security in childhood, significant primary relationships, economic security,
physical security, education, safe birth control and child-bearing4 . Being satisfiers of ba-
sic needs, they are universal satisfiers, valid for all cultures. Universal satisfiers (or inter-
mediate needs), on their turn, are seen as goals with their own specific satisfiers, which
are the means to achieve them.
Also Doyal and Gough see needs, differently from wants, as objective: ‘Not only is it
the case that you can need something that you do not want. You may need it without even
knowing of its existence!’ [8, p. 42]. This is also interesting, as it implicitly introduces
a distinction between needs that are objective (or externally ascribed) and the subjective
perception of needs5 .
After this brief presentation of the state of the art concerning the main notions in-
volved in the conceptualization of needs, we will now turn to a proposal based on some
ontological choices that have to be taken to enable an ontological characterization that
could be useful to structure a need-driven representation of services for the PA.
We cannot conclude our overview of needs’ theories without first having introduced
a further distinction, namely the difference that elapses between needs and desires. Wig-
gins and Dermen [10, p. 62] underline that needs are often conceptually confused with
desires, even if they are quite different from each other. Needs and desires don’t overlap
because, for example, someone could desire something without needing it, and she could
desire it in a more or less conscious state. Conversely, someone could need something
without desiring it. In addition, in order to need something, it is not necessary to be con-
scious of it (e.g., people in a coma are not aware of specific treatments and drugs they
need). In light of this, we can assume that needs and desires are different entities. But
what is the relationship between them?
It might be supposed [11] that needs are a sort of universal human requirement and
that they underlie desires, which depend on a personal perspective. However, as McLeod
remarked [4, p, 13], there are also those who think [12, p. 29] that needs are generated by
desires’ goals. A third hypothesis is that needs are complementary to desires, in the sense
that needs involve desires and vice-versa. This idea has been proposed by Castelfranchi
[13, pp. 863-864], who believes that when someone feels a need, at the same time, she
feels a desire for something that can satisfy that need.
3 At p. 155, though, satisfiers are more generally defined as ‘all objects, activities and relationships which
satisfy our basic needs’. It seems likely that a more thorough conceptual analysis could point to a difference
between intermediate needs (or goals) and the resources that are necessary to achieve them.
4 In their book, Doyal and Gough provide also mechanisms to measure need satisfaction and political and
economical strategies to try to satisfy human needs. Although this is of course a relevant contribution of the
book, we won’t touch such dimensions in the current paper, as our main aim is the ontological analysis of the
primitive notions.
5 Such distinction between needs and wants based on the “objectivity” of the former vs. the subjectivity of
the latter is also present in the philosophical literature. See [9, p. 68] and [3, p. 152].
Actually, as stated above, not all needs are experienced. Moreover, the expression
“felt need” is unclear, because we should make a distinction between a need and the
associated experience. So, as McLeod [2, p. 216] affirms, ‘from the supposition that
urges associated with a need are felt, it does not follow that the need itself is felt’.
In any case, Castelfranchi’s theory of needs has the merit of clarifying, from a psy-
chological standpoint, the cognitive schema that underpins the theory of instrumental
needs. According to him [6, p. 56], we must take into account the following elements so
as to understand such notion of need:
• X has a goal p;
• For this goal, q (action or situation) or y (a resource) needed, q/y is a potential
sub-goal;
• X lacks it;
• So, X cannot achieve p.
3. Towards a proposal
The ongoing digitisation process of Public Administration (PA) aims to make public
services more accessible and effective and to increase citizens’ satisfaction with respect
to them. In this perspective, every service is designed so as to meet citizens’ needs.
Therefore, two themes have become a focus of increasing attention for PA: the quality of
public services and the citizen active role.
The locution “citizen active role” means that the citizen is no longer intended as the
mere beneficiary of the service, but rather as an actor contributing to determine the aim
and the ways of services providing. In fact, citizens’ direct perspective on their own needs
represents a relevant source of knowledge in order to evaluate the matching between
citizens’ needs on one hand and needs that are attributed to citizens by PA on the other.
The rationale behind this distinction is in that the whole decision process on whether
to activate a public service and which one is initially triggered either by a request of
the interested citizen, or by a highlighting of some other citizen or institution, or by a
monitoring process put in place by the PA itself. Moreover, the beliefs of citizens about
their current state may differ with respect to the beliefs the PA has on that same state
and also the citizens’ goals rarely if ever correspond exactly to the goals the PA has with
respect to the citizen’s desirable state.
We might call a need expressed in a citizen request for herself self-attributed need,
a need highlighted by another citizen, institution or the PA itself through its monitoring
processes hetero-attributed need6 .
A useful method of investigation used by PA to measure the service quality as per-
ceived by citizens is the so-called “customer satisfaction”, which allows to understand
the self-attributed citizens’ needs (e.g., by means of complaints and suggestions on PA’s
web pages, opinion surveys, etc.). The perceived quality of service depends on citizens’
6 The literature on services for the PA usually distinguishes between “perceived” – Castelfranchi [6] call
them “felt” – and “actual” needs, but we believe that for the PA it is not so central whether the need is actually
perceived or not, as citizens may self-attribute needs either by feeling them, or by coming to know them in
some other way, for instance if an expert tells them. Also, calling needs “actual” implies the assumption of a
strongly realistic perspective, while what the PA is really interested in is whether such needs may be recognised
based on the information which is available to the citizen and on some parameters that define goal states.
needs satisfaction, which is linked to their expectation with regard to a specific service.
If the citizens’ service experience far exceeds their previous expectations, then the sat-
isfaction will be high; conversely, if the service experience disappoints the citizens’ ex-
pectations, then the satisfaction will be low. The deviation (quality gap) between percep-
tions and expectations allows to quantify the service quality and the citizens’ satisfaction
level.
It is reasonable to assume that the quality gap is in part due to another one, which
is the discrepancy between self-attributed needs and hetero-attributed needs. This is an
epistemological gap concerning the difference between first-person and third-person
knowledge about mental states. Obviously, the citizens’ needs satisfaction can be mea-
sured if and only if we are able to identify their needs and expectations. But how can we
succeed in a well-founded citizen satisfaction analysis, if we don’t even know what ‘to
have a need’ exactly means?
For all these reasons, we believe that it is necessary to work toward a representa-
tion of human needs for PA’s services which takes into account the gaps that we have
discussed.
A first choice we will take is to focus on instrumental needs, rather than on absolute,
basic or fundamental needs, as the latter, though probably morally binding for PA, are so
general that they may constitute sort of “missions” or “inspiring values”7 , rather than a
conceptual tool that can help in the management of its decision processes. So, in what
follows we will use the term “need” to talk about instrumental needs.
Given that our final aim is to model the PA’s decision processes that lead to the
activation of services, which are driven by citizens’ needs, in our framework we will
assume a BDI (Belief-Desire-Intention) perspective. We deem this choice as appropriate
for a twofold reason: it is theoretically well-founded (see the seminal works by Bratman
[14], Rao and Georgeff [15] and Cohen and Levesque [16]) and it has been implemented
in many widely used architectures (JAM8 , Jason9 , SPARK10 among many others), so
that tools and applications are already available for implementation.
Modelling the whole life-cycle of a service would obviously require to model also
what happens “in the world”, including the events that trigger the starting of a PA’s
decision process and how the service that the PA decides to deliver is realized by actual
actions, but in this paper we will focus only on the internal processes, starting from when
the PA becomes aware of a certain situation on which it decides to act and ending when
the PA decides to activate a certain service to respond to a citizen’s need that has been
manifested.
In brief, in the philosophical literature, beliefs, desires and intentions may be seen
as attitudes an agent entertains with respect to a certain representation. Similarly as in
[17], we see mental (or propositional) attitudes – as beliefs, desires and intentions – as
7 What we have in mind here is similar to what Doyal and Gough [8] call “preconditions”, what makes it
possible for human beings to have goals.
8 http://www.marcush.net/IRS/irs_downloads.html
9 http://jason.sourceforge.net/wp/
10 http://www.ai.sri.com/ spark/
~
a mental state an agent is in and that is about11 a certain mental object, whose content
could be expressed by a statement that can be true or false (where “mental object” just
means ‘representation of something else’).
Along these lines, beliefs are informational states of the agent about states of affairs,
general laws, rules, norms, but also other mental attitudes, of the agent herself or of
other agents. Desires are instead motivational states, which are about states that the agent
would like to reach and they can also be about other mental states, of the agent herself
or of other agents. Intentions are deliberative states, in a sense they are desires that the
agent has committed to get realized.
It is worth noting that there are other two fundamental notions in classic BDI ap-
proaches, that of goal and that of plan. Goals are also motivational states and, similarly
to desires, they are states the agent would like to see realized but, differently from desires,
the agent selects them as states to be pursued or brought about, thus they have additional
constraints, like for instance the fact that they have to be consistent with other goals the
agents has selected and this is not necessarily the case for desires. In this sense, it is
more correct to say that the commitment of intentions is directed towards goals rather
than desires. Finally, plans are deliberative states, very similar to intentions, but more
articulated; roughly, they can be seen as concatenations of intentions or, in Bratman’s
words [14, p. 86]:
Plans, as I shall understand them, are mental states involving an appropriate sort of
commitment to action: I have a plan to A only if it is true of me that I plan to A.
Plans, so understood, are intentions writ large.
So, the picture we have so far seems to be the following: an agent has a set of beliefs
(what she believes to obtain in the world, in her mind, or in others’ minds). She also has a
set of desires, states she wants to see realized and she may decide to select, among these
states, those that she wants to pursue (her goals). After the selection, she can commit to
bring about such states and form an intention. Intentions are then expanded and detailed
in plans.
Another kind of informational states are expectations, which can be seen as sort of
beliefs directed towards a future state and, more specifically, to a state that is a goal of
the agent, possibly a goal whose achievement she delegates to some other agent.
So far so good. Now let’s see whether the notions we have analyzed in the previous
section are connected with the ones just characterized and how. Let’s start with needs and
focus on an instrumentalist vision of needs. According to the instrumentalist, sentences
expressing needs have the form: ‘A needs X for P’, where A is an agent and P is a goal
state.
Now, let’s assume that needs are mental attitudes, like beliefs, desires, intentions,
goals and plans. Are they informational, motivational or deliberative states? What are
they about? What are they ontologically dependent on?
11 We are referring here to the notion of aboutness illustrated in [18]. Crane defines intentional objects as
those upon which intentional states are directed. He rejects the thesis that all intentional states have a propo-
sitional content [18, p.31] and holds an internalist position w.r.t. intentional objects and contents [18, p.118].
Due to space constraints, we won’t enter in this paper in the internalism/externalism dispute and postpone a
detailed analysis of mental content to future work.
We agree with Castelfranchi [6] that needs can be seen as instrumental goals12 ,
as without what is needed, what is aimed at cannot be reached or, in other words, we
must obtain what we need before (and for) obtaining what we aim at. For this reason,
we believe needs are motivational attitudes. But let us remind that a need should be
distinguished from what is needed (analogously, a desire should be distinguished from
what is desired, an intention from what is intended etc.). The philosophical literature
on needs distinguishes, in fact, between needs and satisfiers, being satisfiers the states
that satisfy the needs. A distinction can thus also be traced between desires and needs
because, while the former are about states that the agent would like to obtain (and goals
are about desired states the agent selects to be pursued), needs are about satisfiers, states
that, whenever they obtain, make a state the agent aims at (state the agent has a goal
about) obtain13 .
Another interesting aspect of needs is that they also seem to involve a choice (like
goals); intuitively, an agent who has already selected a goal (state) and formed an in-
tention towards it, has to decide what to do (or what should obtain) to reach it. But we
can imagine that at any moment, at least for some goals, there are multiple ways to try
and achieve them; in other words, there are multiple satisfiers for the same goal. When
the agent says (or thinks) ‘I need X to P’, she chooses which of the satisfiers to try and
achieve in order to then achieve P. Needs seem then to depend on goals and on beliefs
(about the best satisfier for that specific goal).
Furthermore, we should be careful not to confuse two different aspects of needs’
conceptual analysis, namely the ontological and the epistemological one. From an on-
tological point of view, as has just been said, needs are motivational attitudes. Episte-
mologically speaking, needs are instead the outcome of an inferential process (based on
the agent’s beliefs) which involves making a choice between alternative satisfiers with
regard to a certain goal to reach. Once the choice is made, the “need for X” becomes an
instrumental goal related to an end goal P.
We can now have a look to the wider BDI general picture, also including needs and
satisfiers. In the previous description, we were left with an intention to achieve a goal
and with the start of the planning process, which consists in detailing the intention. Once
the agent has committed to the execution of the (still partial) plan, she has to take into
consideration some possible ways to pursue the goal (the alternative satisfiers) and to
choose one among them. Such choice amounts to say what she needs in order to achieve
the selected goal. At this point the planning process can go on, as a new intention is
directed towards the specifically chosen satisfier, which becomes now a goal (or, more
precisely, an aimed state) on its own. The agent can now form a new intention about the
state that was previously a satisfier and that has become a goal in the new context. The
process can now be iterated as much as necessary, until the agent realizes (believes) to
have all she needs: the plan is complete and it is now time for action.
A simple example could help here. Gina has a desire to go to the seaside. She be-
lieves, since she has read the weather forecasts on the internet, that on Saturday it will
12 We prefer the term “instrumental goal” to “sub-goal” or “means”, as we would really like to stress the fact
that they help structuring the functional steps in a decision process.
13 This is another reason why we deem the definition of needs as “instrumental goals” particularly appropri-
ate: they are about states that are satisfiers w.r.t. the end goal and that become themselves states an intention
can be about (and so also goals, although instrumental).
be sunny in Villasimius14 and that she will be free and going to the seaside it’s the best
thing she can do for herself. The desire becomes thus a goal that she adopts. She then
starts planning, first forming the intention to go to Villasimius. She begins to reason on
possible ways to reach Villasimius and two come to her mind: either by taking the bus or
by taking the car. She believes that going by bus will take too long and she won’t fully
enjoy the day. She thus chooses, among the possible satisfiers, the one she believes to be
the best according to some perspective. She needs to take the car to go to Villasimius.
Taking the car becomes now a goal and then an intention. But she believes not to own a
car (and not even a driving license!). Among the possible satisfiers, there is either rent-
ing a car with a driver or asking some friends (with a car and a driving license) to go
together. Since the former would be too expensive and less funny, she chooses the latter.
She needs to ask a friend to take the car. And so on, she will choose which friend to ask
and then she will be ready to put her plan into action.
Since our final aim is to apply this analysis of needs in information systems that
could help the PA to structure its service offering on the basis of citizens’ needs, let us
consider, for explanatory purposes, a simple example in the domain of social services,
starting with a citizen’s self-attributed need.
Maria is a single woman with two young children and she is going to undergo a
heart surgical operation. Maria believes that she will be hospitalized for a month and
that, during this period, she will not be able to take care of her children and that there
are no friends or relatives that can do it for her, as she has just moved to a new place, far
away from where she used to live and it would be too much of a burden to ask anyone.
Obviously, Maria desires the wellness of her children, she wants that they are fit, well-
fed and keep going to school throughout her rehab period. Reasonably, Maria’s desire
becomes an adopted goal to achieve.
Notice that from a BDI perspective, in order to achieve a goal, it is necessary (yet
not sufficient) to have a commitment toward the goal, namely, it is necessary to have
an intention. Therefore, suppose that Maria has an intention which is about the above-
mentioned goal. Once Maria has formed the intention, a means-end problem is posed:
how will her children be taken care of?
At this point, Maria starts thinking she needs someone who will able to look after her
children, and she reasons about possible satisfiers and, based on her beliefs, she makes a
choice, which is to ask for the help of a social worker to social services. Now, the need
for a social worker becomes an instrumental goal with respect to Maria’s end goal and,
as soon as an intention is formed, she will be ready to act and go asking to PA15 .
So far we have talked about Maria’s self-attributed need but, given that Maria cannot
satisfy her need by herself, this must be expressed to someone so as to obtain the related
end goal16 . Therefore, imagine that Maria proceeds to the social services departments for
14 A place with wonderful beaches in Sardinia.
15 Obviously, the process of means-end reasoning must be means-end coherent. For example, Maria can’t
include in her plan the intention to ask her husband to look after the children, because she is a single woman
and the means in this case are not coherent with her beliefs about her current state.
16 It must be emphasized that, ontologically speaking, the expressed need is different from the need itself.
The former is a triggering event (as defined in [19]), an input which produces an entry into the service’s system
and alerts the PA to take charge of some citizens’ needs, while the latter is in our framework a mental state. A
need can be expressed also on the basis of an “hetero-attributed need”, which means that it can be highlighted
by another citizen, institution or the PA itself through its monitoring processes; moreover, it is not necessary
that the agent who expresses the need and the one who first attributes it coincide. For instance, a teacher could
a preliminary interview with a PA’s employee, in which she has the opportunity to express
her need. We can suppose that the employee listens carefully to Maria’s expressed need
and that he asks for clarifications about it. Then, basically, two alternative cases may
occur: (i) The employee attributes the need to Maria; (ii) The employee does not attribute
the need to Maria.
Regarding (i), we can assume that Maria’s expressed need falls within the em-
ployee’s competence and that he has previously checked that Maria meets the specific
requirements for the activation of the social assistance service. Therefore, in this case,
the employee will proceed with service activation.
With respect to (ii), the employee could believe that Maria’s need is not within his
remit or he may not be able, for some reason, to identify or understand the need in
question. In this case, we can suppose that Maria’s expressed need won’t be satisfied
(in the way she expected and has requested), but her data will be anyway registered and
sent to some competent office that maybe could attribute a different need (a different
instrumental goal to be satisfied in order for her children to be taken care of) and activate
a related service.
However, in both cases, if Maria does not meet the specific requirements for the
activation of the social service which is related to her need, the employee will not proceed
with service activation. The latter is an institutional fact which concerns laws, rights,
and rules that are, from the ontological standpoint, social objects, so they fall outside the
scope of the “internal” perspective we are focusing on in this paper.
4. Related works and main contributions
To the best of our knowledge, few ontological frameworks trying to model needs’ driven
services have been proposed so far. Exceptions worth to be mentioned are [20], [21]
and [22]. Though with different aims, these works try to model the whole service’s life-
cycle and assume an already available theory of needs ([20] builds on Kotler’s theory
[23], centered on marketing, [21] on Maslow’s hierarchy of needs [7], and [22] on Max-
Neef’s works [1]). They are thus more interested in classifying specific kinds of needs
in a taxonomy and connecting them to services, but they all take the concept of need
as given, without providing a foundational analysis of it. Hence, the main contribution
of this paper in the debate is towards a deeper understanding of the concept of need, so
as to better specify its semantics before using it in a service system representation. In
addition, we have faced the issue of the epistemological gap created by the difference
between first-person knowledge and third person-knowledge about needs, proposing a
conceptual distinction between self-attributed and hetero-attributed need, which could
be helpful in accounting for citizens’ satisfaction and expectation in the design of public
services. Future work shall be dedicated to provide a logical account of our conceptual
analysis of needs, their properties and relations with other mental attitudes. Moreover, it
could be interesting to analyse also the concept of expectation and its connection with
goal frustration.
attribute the need for psychological assistance to a student who seems to be depressed; she could then express
such need to some authority or talk with the student who, after having self-attributed the need to himself on
the basis of a hetero-attribution, decides to express it. Moreover, it is worth noting that both for self- and
hetero-attributed need, the attribution could be wrong.
References
[1] Manfred Max-Neef. Human Scale Development. Apex Press, New York, 1991.
[2] Stephen K. McLeod. Knowledge of need. International Journal of Philosophical Studies, 19(2):211–
230, 2011.
[3] David Wiggins. Claims of need. In David Wiggins, editor, Needs, Values, Truth, pages 1—57. Blackwell,
Oxford, 2nd edition, 1991.
[4] Stephen K. McLeod. Two philosophies of needs. Polish Journal of Philosophy, 9(1):33–50, 2015.
[5] Guy Fletcher. Needing and necessity. In Mark Timmons, editor, Oxford Studies in Normative Ethics:
volume 8, pages 170–192. Oxford University Press, 2018.
[6] Cristiano Castelfranchi. To believe and to feel: The case of “needs”. Technical Report FS-98-03, AAAI,
1998.
[7] A. H. Maslow. A theory of human motivation. Psychological Review, 50:370–396, 1943.
[8] Len Doyal and Ian Gough. A theory of human need. Critical perspectives. Macmillan, 1991.
[9] G. E. M. Anscombe. Intention. Harvard University Press, 1957.
[10] David Wiggins and Sira Dermen. Needs, need, needing. Journal of medical ethics, 13:62–68, 1987.
[11] Soran Reader. Aristotle on necessities and needs. Royal Institute of Philosophy Supplement, 57:113–
135, 2005.
[12] Raymond Plant, Peter Taylor-Gooby, and Anthony Lesser, editors. Political Philosophy and Social Wel-
fare: Essays on the Normative Basis of Welfare Provisions, volume 19 of Routledge Revivals. Routledge,
London, 1980.
[13] Cristiano Castelfranchi. Goals the true center of cognition. In Fabio Paglieri, Luca Tummolini, Rino
Falcone, and Maria Miceli, editors, The Goals of Cognition. Essays in Honor of Cristiano Castelfranchi,
pages 825–870. College Publications, London, 2012.
[14] M. Bratman. Intention, plans, and practical reason. Harvard University Press, Cambridge, MA, 1987.
[15] A. S. Rao and M. P. Georgeff. Modeling rational agents within a BDI-architecture. In Principles of
Knowledge Representation and Reasoning. Proceedings of the second International Conference, pages
473–484, San Mateo, 1991. Morgan Kaufmann.
[16] Philip R. Cohen and Hector J. Levesque. Intention = choice + commitment. In Proceedings of the Sixth
National Conference on Artificial Intelligence - Volume 2, AAAI’87, pages 410–415. AAAI Press, 1987.
[17] Roberta Ferrario and Alessandro Oltramari. Towards a computational ontology of mind. In Achille C.
Varzi and Laure Vieu, editors, Formal Ontology in Information Systems, Proceedigs of the Intl. Conf.
FOIS 2004, pages 287–297. IOS Press, 2004.
[18] Tim Crane. Elements of Mind: An Introduction to the Philosophy of Mind. Oxford: Oxford University
Press, 2001.
[19] Roberta Ferrario and Nicola Guarino. Towards an ontological foundation for services science. In
Domingue et al. [25], pages 152–169.
[20] Sybren De Kinderen, Pieter De Leenheer, Jaap Gordijn, Hans Akkermans, Rose-Marie Dröes, and
Franka Meiland. e3 service: An ontology for needs-driven real-world service bundling in a multi-
supplier setting. Applied Ontology, 8(4):195–229, 2013.
[21] Bart Gajderowicz, Mark S. Fox, and Michael Grüninger. Ontology of social service needs: Perspective
of a cognitive agent. In Ludger Jansen, Daniele P. Radicioni, and Dagmar Gromann, editors, JOWO 2018
– Proceedings of the Joint Ontology Workshops 2018, volume 2205. CEUR Workshop Proceedings,
2018.
[22] Shawn Dexter Dsouza. Cloud-based Ontology Solution for Conceptualizing Human Needs. PhD thesis,
University Of Ottawa, 2015.
[23] P. Kotler. Marketing Management. Pearson One. Pearson Prentice Hall, 2009.
[24] Michael E. Bratman, David J. Israel, and Martha E. Pollack. Plans and resource-bounded practical
reasoning, 1988.
[25] John Domingue, Dieter Fensel, and Paolo Traverso, editors. Future Internet - FIS 2008, First Future In-
ternet Symposium, FIS 2008, Vienna, Austria, September 29-30, 2008, Revised Selected Papers, volume
5468 of Lecture Notes in Computer Science. Springer, 2009.
[26] Fabio Paglieri, Luca Tummolini, Rino Falcone, and Maria Miceli, editors. The Goals of Cognition.
Essays in Honor of Cristiano Castelfranchi, volume 20 of Tributes. College Publications, London,
December 2012.
[27] Maria Miceli and Cristiano Castelfranchi. Irrevocable goals. Review of General Psychology, 21(1):69–
81, 2017.