facilitatorsamplifiers for institutional agenda setting

desertdysfunctionalInternet and Web Development

Dec 4, 2013 (3 years and 11 months ago)

116 views

S
OCIAL MEDIA FOR MUSE
UMS
.

A COMPARATIVE ANALYS
IS OF WEB COMMUNICAT
ION AND THE
BLOGOSPHERE

ABSTRACT

Blogosphere is emerging as a strategic dimension of public debate and can be considered
as a social space of communication where discourses
related to cultur
al heritage could be
tracked and analysed. According to Henry Jenkins, blogs have the role of
facilitators

for public
discourses: they basically work as
amplifiers for institutional agenda setting

(Jenkins, 2008).
Starting from this consideration, our pap
er tries to answer the following questions: what kind
of discourses would emerge from the blogosphere? Which topics or images are debated by
bloggers and could we compare them with those disseminated by museums web sites?

Our paper investigated those rese
arch questions through content analysis and cluster
analysis supported by the prototype software CoOperare, and it revealed a controversial the
role of the
blogosphere
compared to that of the institutional communicatio


1. Introduction and research questi
ons


Among the most popular web 2.0 expressions, blogs enhance communitarian, spontaneous and
basically self
-
regulated communication processes. The Blogpulse portal, a
constantly updated
observatory about the state of the art, counts today around 147 mill
ion blogs on the net
1
. Considered
the increasing popularity of this new communication environment, it is possible to identify a
blogosphere
:
a new dimension for the public sphere, a communication space where it is possible to define and
analyse the emergen
t social issues and opinions. Blogs represent new fields for the social research
aimed at understanding “social discourses” in their “becoming media” and confronting them with
mainstream media. Usually blogs are identified as new forms of “participatory
journalism”: a “peer
journalism” managed by individuals or groups with different political, ethical or intellectual positions.

O
pinion leaders’ blogs such as Salam Pax
2

or Instapundit
3

or the Italian Beppe Grillo
4

are very popular
as well as their controve
rsial positions if compared to the agendas of traditional mass media:
newspapers, TV, etc. Besides the so
-
called “blog
-
stars”, the World Wide Web counts millions of blogs
posted everyday by common users on various subjects: sport, fashion, gossip, tv, etc.
, or even related to
personal biographies and experiences.

According to Henry Jenkins (Jenkins, 2008),

blogs are “facilitators” of public discourses:




they basically work as amplifiers for institutional agenda setting
"

. Blogs are not necessarily t
o be considered
as
opposition media
, but they often play the role of amplifiers for mainstream issues. The empirical
research on journalism blogs took in careful consideration the dialectic between mainstream media
discourses and users generated contents.
Many researches compared how specific public issues and
general themes such as politics, religion, etc. have been argued by blogs and by mass media at the same
time (Campbell 2010; Xenos 2008; Totaro 2008; Colombo, Murru 2007). Other works concentrated on
the motivations issue related to journalism blogs authors (Trammel et. al., 2006); some other researches



1


http://www.blogpulse.com/

2


http://salampax.wordpress.com/

3


http://pajam
asmedia.com/instapundit/

4


http://www.beppegrillo.it/

focused on the analysis of the relation between mainstream media agenda setting and that arising from
user generated blog
-
post (Meraz 2009; Kenix 2009)
.

In the cultural heritage domain, even
though social media relevance was recognised in the academic
debate (Russo et al. 2006), the empirical investigation still focuses on institutional blogs field. There are
a number of museum web sites, especially in
US, which present participatory tools such as blogs, social
networks, You Tube or Flickr channels. Researches in this field tried to prove how social media could
enhance the institutional communication in terms of dissemination, education or in the devel
opment of
“brand communities” (Liu et al., 2010; Trabill et. al. 2009; Samis 2008). According to this research
scenario, our contribution highlights a new topic, trying to analyse the blogosphere considered as a
“spread public sphere” in comparison with th
e institutional communication web tools.

In recent years museums’ blogosphere significantly enlarged: the first informal survey in March
2006 counted around 30 blog
s on a total of 30 million blogs on the web. At the beginning of 2007,
MuseumBlogs.org
5
, the

most popular portal in the field, listed more than 100 blogs (Chan 2007), while
today around 400 blogs are counted, including the institutional ones and those managed by individuals.
Those latter kind of blogs are particularly significant in order to iden
tify the social discourses about
institutions, and their analysis can answer to the following questions: which are the emerging issues and
the social representations? How blog
-
discourses correspond to the representations spread by
institutional media?
6



2. The institutional communication: virtual museums

When analysing the institutional communication, our research focused on official web sites
considered as mainstream sources. Differently from other institutions, Italian museums have a poor
relevance in
mass media (newspapers, television, radio etc.): they just appear
thanks to their events or
in special occasions such as public anniversaries and celebrations (e.g. “Open Museums Day”, once a
year). Due to this low visibility in mainstream media, we assum
ed that museums institutional
communication takes place preferably through museums official web sites. The web provides museums
with an extraordinary opportunity to enhance their promotion policies. Virtual museums first appeared
in the mid’ nineties and
their number increased following the development of the World Wide Web as
a global medium of communication. A number of distinctive features emerged thanks to the first
studies in the field: a lot of museum websites appeared like electronic brochures refle
cting their
corporate images, while some others exploited the most innovative and interactive communication
potentials of the web (de Meijere, 1996). At the beginning of 90’s just a few museum web sites
presented forum and community applications, even tho
ugh an horizontal and grass
-
root
communication development seemed already at those time the most hopeful perspective (Micelli et al.
1998).

Virtual museums represent a strategic opportunity for two main reasons:





t
he potential for popular and new institu
tions to promote themselves with low costs of access
and the opportunity to reach, at the same time, a very large audience. Compared to expensive
virtual reconstruction and simulation projects, web communication provides even the small
museums, foundation
s and archives with the opportunity to develop an autonomous visibility
online space with no expense and significant profits in terms of promotion and marketing;



t
he spread of the virtual museums follows the general web audience increasing: between 1998
an
d 2000, the number of users connecting at least once a month raised from three to eighteen
million of Italian, involving almost the 30% of the entire population (Pasquali, Scifo 2004). A
relevant audience of users and virtual visitors emerges, a
critical m
ass

using the net ever more as
an essential tool for information and entertainment. Websites besides, become strategic tools



5


http://www.museumblogs.org/

6

Our research refers to the project PRIN (Research Project of National Interest) CoOperare (
CO
NTENT
O
RGANIZATION
,

P
ROPAGATION
,

E
VALUATION AND
R
EUSE THROUGH
A
CTIVE
RE
POSI
TORIES
)

coordinated by Prof. Mario Ricciardi, Turin Politecnico.
Further information about
CoOperare project are to be found at
http://nexos.cisi.unito.it/joomla/cooperare/
.

for cultural tourism: they give a quick and easy access to resources for the organization of a trip,
a museum visit, a convention,

etc.

The web is a central phenomenon for the cultural heritage dissemination: that is proved by the
increasing number of on
-
line institutions and projects related to the heritage digitization and
connection through the communication networks (Capaldi et
al. 2008; Granelli, Traclò 2006).

The virtual museums’ spread in Italy follow
ed the international trend even though nowadays it is
possible to identify relevant differences especially in the institutional use of participatory media such as
blogs and soci
al networks. According to University of Urbino’s research, there are a number of
European museums who make a large use of social tools, like Facebook, YouTube or blog portals such
as Wordpress: among the best practices, a particular relevance deserve Germa
n and English museums:
they’re particularly committed in creating and managing communities, blogs, video

blogs etc.
(Capaldi
et al.,
ib.
).

Italian museums still use the web in a very traditional and “static” way.

According to this consideration, the blo
gosphere could actually represent an alternative source and
a great potential to enrich public and off
-
stream discourses.



3.
The design of the research: the “brand name” communication


In defi
ning the significant sample of Italian virtual museums,
our research privileged the
brand name

domain. This term describes institutions which represent the principal destinations for Italian cultural
tourism. This phenomenon is particularly relevant in this country, where a small number of museums,
artistic and

archaeological sites, attracts a visitors audience which is much wider than the one of other
cultural sites: in fact, the first 30 Italian museums, which represent the 8% of the total (around 400),
attract the 75% of all visitors, around 25 million of peo
ple: among those 30, the first 9 involve 17
million of visitors (Antinucci 2007). Data reveal then a strongly polarized scenario focused on a small
number of “museum
-
stars”: those museums represent the sample
-
field where it is possible to identify a
core o
f communication strategies and public discourses which involve both the institutional
communication and the one generated in the blogosphere participatory environments. According with
official statistical resources
7
, we identified 15
brand name

representi
ng the Italian institutions (museums,
artistic or historical sites, environment centres) that attract more than 500.000 visitors per year (see the
table below).






7


Among t
he sources considered: the SISTAN (Sistema di rilevazione statistica del Ministero per i Beni e le Attività
Culturali) reveals exclusively data related to Public Institutions; the Finanza e Turismo portal, an information tool realise
d by a
collaboration am
ong Assoturismo, Confturismo, Federturismo, Confindustria monitor both public and private institutions; the
Touring Club Italia which is the most exhaustive and up
-
to date source related to public, private and mixed institutions such as
foundations etc...


Table 1: The 15
brand name

of Italian cultural heritage. Source: TCI Research Centre (2008
) (eds.),
Museum Dossier 2008,
http://www.touringclub.it/ricerca/dossier_index.asp?area=dossier
, September 2010.


This sample represented the
analysis field for the institutio
nal communication survey: each
institution has, in fact, an autonomous web site or a dedicated section in a portal (e.g. the Uffizi Gallery
is part of the City of Florence museums or the Capitolini Museums appears in the City of Rome
museums portal). Web s
ites present different contents related to both communication topics and
secondary information such as the practical information about museum location, tickets reservation,
guided tours etc.. In the first phase institutional web sites have been analysed th
rough a textual analysis
aimed at finding the more significant communication web pages: the pages related to the heritage
dissemination, the ones describing the institution’s collections and history, those about events and
temporary exhibitions including a
lso the events archived, the press releases concerning arts and
educational activities as well the pages related to museum’s staff and the editorial activities, have been
included in the textual sample considered for the analysis.

The survey’
s outcome was

then an homogeneous corpus of documents which represent the
brand
name

institutional communication.

4. Research methodologies:
comparison between institutional and blogosphere lexicons.

The blogosphere is a new field of investigation for the analysis of
social discourses: a hybrid
“genre” that combines the journalistic writing style, the diary, the informal interview, etc
. (Antelmi
2006). In general, blogs represent a set of persistent, searchable, replicable and scalable contents (Boyd
2007), while from
a more specifically sociological point of view blogs offer spontaneously user
-
generated contents that can be qualitative and quantitative analyzed.

In addition, the amount of available contents is becoming a very important issue: the blogosphere
represents

a huge volume of stored data always available for analysis with tools such as search engines,
text indexing applications and lexical analysis tools like the one proposed in the “Cooperare” project.
Considering the increasing volume of data produced by the

blogosphere, it becomes strategically
important to develop novel methods and tools in order to identify the most representative contents. In
fact, many institutions are using "dedicated" search engines to retrieve a first selection of contents:
among the
others, some of the most famous are the Technorati blog search engine, Google Blog
Search and BlogPulse (Carey, Jeffrey 2006). But even considering these systems, it is essential to first
provide a more refined tool to further limit the volume of data (for

example, the query "Musei
Vaticani" results in Google Blog Search with more than 10,600 italian results) and, secondly, to reduce
the overall redundancy and the "noise" related to the irrelevant information (such as tags, advertising,
etc). According to t
his objective, it has been developed the “Cooperare” application, a system that is
able of selecting, according to user defined parameters, a corpus of text (blog posts) relevant to the
institutional brand name. This tool has been then integrated with the

search engine
Google Blog
Search, which provides a selection of recent posts based on a user defined query (in our case the names
of the institutions concerned). In fact, Google Blog Search combines the latest techniques within the
field of information re
trieval with the solidity of the most used search engine on internet. Then, given
the set of the first 1000 blog posts for each considered institution (written from January 1999 to April
2008), we applied standard information retrieval techniques in extrac
tion of document surrogates that
use a vector representation (Salton, Buckley 1988) for each post that appears in the corpus. The same
procedure has been then performed on the set of documents extracted from the institutional web
pages. Given this informat
ion formalization approach, the main objective of the application is the
lexical comparison between the set of posts from the blogosphere (and, thus, representing the point of
view of visitors with respect to the considered institutions) and the whole docu
ments extracted from
the official websites of the institutions (that define the image they proposed about themselves to the
potential visitors). Assuming that each term used in web pages is strongly charactering the institutional
communication, the compari
son with the lexicon of the blogosphere provides user
-
generated
quantitative evidence of the following properties:




the lexicon used by both sources ;



the level of correspondence between the two lexicons;



correspondence between the terms that characterize
the two lexicons.


More specifically, we defined two metrics in order to better highlight the degree of overlap between
the used vocabularies. The first one defines the intersection between the two vocabularies, calculating
the number of terms shared by tw
o sources. The second one aims to calculate the distance vector that
would be obtained considering the two sets as term
-
vectors; in order to perform this operation, we
define the set of words shared by the two sources as a vector, where each dimension repr
esent the term
frequency. Then, for each considered institution, we calculate the cosine similarity measure among the
vectors; the result represents a "mathematical" measure of the degree of overlap between the two
vocabularies and reports the numerical d
egree of similarity that otherwise would remain uncertain and
unquantifiable. In addition to these information, the application presents, for each institution, three
additional information:



shared vocabulary: this represents the set of terms shared by the
blogosphere and institutional
web sites (ordered by frequency). It is also possible to choose the order with respect to the set
of documents in the blogosphere or institutional web sites;



not shared vocabulary: this represents the set of terms, reported b
y the institutional web sites
that are not shared by the users of the blogosphere (sorted by frequency value);



shared tag cloud: it represents the shared set of terms that better characterizes the blogosphere
and information included within the institution
al web sites.


These three information provide an overview of the vocabularies expressed by users and museum.
Figures 1 and 2 show the results for the “Museo Nazionale del Cinema” and the “Musei Vaticani”.



Figure 1: Results related to the “Museo Naziona
le del Cinema” provided by the
“CoOperare” tool.


Figure 2: Results related to the “Musei Vaticani” provided by the “CoOperare” tool.

4. 1 Lexicon of the museums: preliminary results and quantitative interpretations

The preliminary results provide a quan
titative estimate of the similarity between the used vocabularies.
The considered lexicons only contain nouns; the adjectives, the verbs, the adverbs, etc. have been
excluded
by the analysis process. Figure 3 shows, for each museum, how the most important
terms
reported in the institutional web sites are also referred in the blogosphere.

61%
43%
52%
56%
44%
48%
50%
46%
58%
54%
46%
47%
57%
50%
63%
0%
10%
20%
30%
40%
50%
60%
70%
Musei Vaticani
Museo di San Marco
Galleria Borghese
Musei Capitolini
Bioparco di Roma
Museo delle Antichità Egizie
Museo Nazionale del Cinema
Acquario di Genova
Galleria degli Uffizi
Galleria dell'Accademia
Museo Centrale del Risorgimento
Museo Nazionale di Castel S. Angelo
Opera di Santa Croce
Palazzo Ducale di Venezia
Scavi di Pompei

Figure 3: Shared vocabulary between institutional web sites and the blogosphere.

It is interesting to analyze the results provid
e
d by the comparison between the “not shared”
vocabulary and the “shared” one. In this case, the reference is the institutional lexicon that can be
considered as "noise"
-
free information with respect to the blogosphere, which often includes irrelevant
data

(such as dates, names or nicknames of the bloggers, advertising, code snippets, HTML tags, etc).
Moreover, the blogs also include colloquialisms, typical of the spoken discourses (slang expressions,
etc). However, the content of the virtual museums web
-
si
tes are essentially textual in nature and present
richer vocabulary and higher linguistic correctness. Compared to the institutional vocabulary, it is
interesting to note that the blogosphere is characterized by the absence of specialized terms of proper
n
ames (e.g. Bernini, Praxiteles, Cabiria, etc). Looking at the values that the application assigns to each
term that has been reported by the blogosphere, it is clearly evident that these weights are rather low
(for example, the term “interconfessionale” ha
s a value of 0.048 with respect to a maximum possible
value of 1). This behavior highlights that even within the institutional web sites this term does not
appear so frequently. The institutional communication is then characterized by an overall lexicon
si
mplicity, but can still be considered very different from the writing style of the blogosphere. In fact,
the blogosphere fundamentally uses a very informal and non
-
specialist language. Moreover, the
discourses expressed by the blogosphere are not related t
o the specificities of the museums and this
obviously explains the absence of terms such as painting, sculpture, etc.


4.2
Evidences from the analysis of the shared tag cloud


The tag cloud shared by the blogosphere and museums web sites adds significan
t information to the
quantitative analysis of the lexicon; in fact, the tag cloud includes only terms that occur more frequently
(the first seven), while the overall comparison between vocabularies consider all the terms reported by
both sources (ordered w
ith respect to the frequency). Moreover, it permits a comparison between the
terms that appear more frequently within the institutional web sites with those more representative of
the blogosphere.

The overlapping permits to identify a group of 7 terms whic
h report the terms that are more frequently
used by both the blogosphere and the institutional web sites. The terms are also visually represented
with different font sizes depending their importance in the considered sources.

Figure 4

illustrates the tag c
loud of the “Galleria degli Uffizi”

The terms "mostra", "galleria", etc.. are more characteristic than the terms "artista", "storia" and
"palazzo" (visually shown with smaller font sizes).





Figure 4: Shared tag cloud related to the “Galleria degli Uffi
zi”.

It is important to remember that the tag cloud is commonly used in the web to easily represent the
most popular terms within the considered web page: web portals and search engines generally report
the most searched terms in the web.

In our case, the

tag cloud does not refer to a single source (a newspaper or a search engine) but
represent the result of the combination of institutional sources and selected blog posts.

In order to interpret the tag cloud, we performed a two phases analysis: within the

first step, through an
inductive process, we tried to identify similarities between the terms reported by the tag clouds in order
to identify high level categories of the considered institutional museum.

In the second phase we have performed an empirical
evaluation of this group of categories by
performing a clustering operation on the data (i.e., we automatically retrieved the group of institutional
museums based on the terms they expressed in their web pages). These two
phases are described in the
next
sections in detail.





4.2.1

First step: an interpretative analysis of the shared tag cloud


Taking into account the terms that appear in the shared tag cloud, we inferred the categorizations that,
in our opinion, best describe the considered museums and

specificities of each of them.
In the first
category, "institutions that organize exhibitions", has been grouped the following institutions: the
“Musei Vaticani”, “Galleria Borghese”, “Musei Capitolini”, “Galleria degli Uffizi”, “Galleria
dell’Accademia”,

“Museo Nazionale di Castel S. Angelo”, “Palazzo Ducale” in Venice, “Museo
Centrale del Risorgimento”, “Scavi Archeologici di Pompei”, “Museo Nazionale del Cinema”.
Among
these, seven institutions explicitly reported, among the most important terms in the
related tag cloud,
the term "exhibition". The web sites of these institutions are very rich information containers, and
provide depth and specific knowledge about various aspects of the considered domains. Nevertheless,
the issues on which the blogosphere
and institutional communications are converging is represented by
very popular current events: exhibitions in fact represent the present, the events that are potentially
more interesting for the users of the Web and all the "potential" visitors. These ins
titutions are
therefore strongly characterized as event venues (see for example the tag cloud of the “Musei Vaticani”,
“Museo Nazionale del Cinema” and the “Galleria degli Uffizi”, reported in the previous pages): we’ll
call them the
event museums
.


Howeve
r, the second category includes institutions that represent historic buildings or places of
worship. This category includes the “Palazzo Ducale” in Venice, the “Museo di San Marco” in Venice
connected to the Basilica, the “Opera di Santa Croce” (Church of
Santa Croce in Florence), the
“Museo Centrale del Risorgimento” in Rome (located within the Vittoriano). Analyzing the tag clouds
related to the institutions represented by this category

(figures 5, 6, 7 and 8), the most characteristic
terms are those that

indicate the location where they are (Venice, Florence, Italy .. etc.) and nouns as
“palazzo”, “chiesa”, etc. These terms denote them as "an historical and artistic buildings." Thus, the
overall evaluation of these characteristic features could be summari
zed, in short, by considering their
nature of "Monuments" or
monument museums
.





Figure 5: Shared tag cloud related to the “Palazzo Ducale” of Venice.




Figure 6: Shared tag cloud related to the “Museo di San Marco”.




Figure 7: Shared tag cloud
related to the “Opera di Santa Croce”.



Figure 8: Shared tag cloud related to the “Museo Centrale del Risorgimento”.

We can also observe that this characteristic has been indentified also with institutions that organize
exhibitions, like the “Museo Nazi
onale di Castel S. Angelo”.

The third and last category is a hybrid one, very hard to be unequivocally described: the Aquarium and
the new “Bioparco” of Rome represent a sort of evolution of the traditional "zoo" and "aquarium"
since today they offer muc
h more to visitors than just visiting and seeing rare animals. They are
entertainment centers offering educational initiatives, study and research opportunities, seminars and
conventions; moreover they can also serve as a venue for conferences, parties or
industry events. They
are in fact multi
-
purpose centers widely different from the old ideas of zoo and aquarium; they tend to
make culture through their “living exhibitions”, stimulating the public debates, promoting projects for
schools, families etc. T
he success in terms of visitors of these centers proves their uniqueness and
defines them as a tourism destination for all kind of purposes. We’ll assume them as
environment centers
.

The analysis of the shared tag clouds revealed that the most popular ter
m for this type of
institution is "environment" and proves that the speeches in the blogosphere and institutional
sources consistently reflect the mission of these institutions.

Figures 9 and 10 show the data related to “Acquario di Genova” and “Bioparco d
i Roma”.


Figure 9: Shared tag cloud of

the “Acquario di Genova”
.









Figure 10: Shared tag cloud of

the “Bioparco di Roma”
.


4.2.2

Second Step: testing the interpretations with cluster analysis

With the objective of verifying the validity of the categor
ies previously identified, we performed a
cluster analysis of the vocabularies of the blogosphere and institutional websites. Within this step, we
separately analyzed the two considered lexicons (in contrast with the previous analysis that only
considered
the shared tag clouds) in order to more precisely analyze the semantic relationships that
exist among the considered institutions. Thus, we first analyzed the vocabulary of the blogosphere and
then the one regarding the institutional communication, highlig
hting the clusters that emerge from this
comparisons. In other words, given a set of institutions, the cluster analysis process separates the
original set into subsets (called clusters) so that two elements (institutions in our case) in the same
cluster ar
e more similar that two elements in two distinct clusters. The clustering approach is a
common technique for statistical data analysis used in many fields, including machine learning, data
mining, pattern recognition, image analysis and bioinformatics.

T
he results of cluster analysis highlight a significant overlap between the retrieved categories and the
ones induced by the previously reported free interpretation, especially with regard to the blogosphere.





Figure 11: Shared tag cloud related to the f
irst category “Historical monument or building”,
retrieved based on the Blogosphere lexicon.




Below, the clusters identified by applying the very well known in literature XMEANS (Pelleg and
Moore 2000) algorithm on the blogosphere:


C
luster 1) Historic
al monuments or builidings


Scavi di Pompei”


Palazzo Ducale di Venezia”


Cluster 2)


Bioparco di Roma”


Acquario di Genova”


Cluster 3)



Galleria degli Uffizi”


Musei Vaticani”


Galleria Borghese”


Museo di San Marco”


Museo Nazionale di Cas
tel S. Angelo”


Opera di Santa Croce”


Museo Nazionale del Cinema”


Accademia di Firenze”


Cluster 4)


Musei Capitolini”


Museo Centrale del Risorgimento”


Museo delle Antichità Egizie”

Comparing these categories with the ones previously identified,
we can notice that there is a significant
overlap: for example, the first category reflects the type
monument museums

(and the related the tag cloud
reports terms such as “Firenze”, “palazzo” etc., see figure 11). The same analogy can be found in the
secon
d category,
environment centers
, and in the third one regarding the institutions organizing temporary
exhibitions,
event museums
.

Figure 12: Shared tag cloud related to the first category retrieved based on the institutional
lexicon.



However, analyzin
g the categories derived from the institutional web sites, there is a significant
homogeneity. The retrieved clusters are only two:


1) Cluster 1


Museo Centrale del Risorgimento


Bioparco di Roma


Acquario di Genova



2)Cluster 2


Musei Cap
itolini


Musei Vaticani


Galleria Borghese


Museo Nazionale del Cinema


Galleria degli Uffizi


Museo di San Marco


Opera di Santa Croce


Museo delle Antichità Egizie


Galleria dell'Accademia


Museo Nazionale di
Castel S. Angelo


Scavi di Pompei


Palazzo Ducale di Venezia


In short, it is possible to observe a lower specificity of the terms of very general categories. One of the
possible explanation of this behavior is the st
andardization of the vocabula
ry: the terms reflect a
standard vocabulary of a typical institutional communication (Figure 12). In fact, we can observe that,
for communication purposes, the web flatten the characteristics of each individual institution.
On the
other hand, it is interes
ting to
note that the blogosphere highlights some distinguishing features and the
specificities of each analyzed institutions; thus, the retrieved categories reflect the real nature of the
considered museums. Therefore, the institutional communication turn
s out to be more “flat, shallow,
etc.” than the one elaborated in the participatory blogs.

5.
Conclusions and further reflections

In this paper we compared the communication of institutional web sites of the most important
Italian museums (brand names) wi
th the corresponding social discourses contained in the blogosphere.

In detail, the cluster

analysis process reveals significant evidences related to two main aspects:

1) the linguistic richness and variety, the style of the discourses, the type of lexico
n used;

2) the contents of the discourses or the emergent topics.

Regarding the first i
ssue, the most homogeneous clusters arise from the institutional
communication: they show in fact an homogeneous lexicon. The institutional communication's
language tur
ns out to be a normalized, quite stereotyped language, which is likely to be a specialized
language typical of the different typology of institutions. That's also what emerges from the lexical
analysis described in par. 4.1. The communication style of the
brand names

appears as more
undifferentiated than the blogosphere's one. When observing the results emerging from blogs, it is
possible to identify in fact four different clusters which significantly overlap those we inferred through
a deductive reasoning
in par 4.2.1.

With respect to the contents we can point out two different outcomes; the first tends to reinforce
Jenkins' hypothesis: blogs give relevance to the topics already presented by the institutional
communication web sites. The clusters retrieved
from the blogosphere reflect in fact the 3 categories
previously described (see the shared tag cloud analysis): the
event museum
, the

monument

museum

and the
museum as an environmental center
. Blogosphere’s contents, though, are quite far from being
agai
nst

or
in
opposition

to institutional discourses, but they rather amplify and re
-
present the mainstream issues.

Our second observation goes even further: It reveals that blogs represent the
brand names

mission
even better than institutional web sites do.
It seems, in fact, that blogosphere emphasizes the 3 Images
previously described with an even richer and more detailed discourses, and this should highlight that
Blogs are more mainstream than official web sites.

How could we explain that ?

In his famous
essay,
L'archéologie du savoir
(Foucault, 1969),

Michel Foucault said that the order of a
discourse is given by an invisible set of options or hidden rules which can shape not only the structure
of the discourse but the content itself: this options set af
fects what is allowed to be written and to be
known. The substance of this order is political, social and it relies upon social relations based on power
and experiences. This hidden order is basically extra
-
linguistic and extra
-
communicative but it
dramat
ically constraints different typologies of discourses related to an object (the object
clinic

for
example, Foucault
ib
.) and produces an extremely complex and intertwined set of messages and
meanings. This hidden order could affect the institutional disco
urses, such as the medical ones, as well
as the popular ones exchanged by common people about diseases and insanity, as well as the political
discourses which have an impact on the administrations policies related to hospitals and the public care
(Foucault

ib
.).

Even if our research is still quite far from identifying an hidden discourse implied in the elaboration
of the idea of
museum
, it seems to us that Foucault reflections may have a great potential in order to get
a deeper understanding of the inner s
tructures of the
museum discourses
. Foucault's statements may help
us not only to identify what is emerging from two
apparently

different sources, the institutional one and
the so called
grass root
communication
, but also to reveal the social process dynam
ics which are implied
in the elaboration and expression of such discourses. Our research results show in fact an interesting
and not so easily predictable similarity between those discourses in terms of contents. Further
researches will be needed in order
to recognise the order implied in those processes and, above all, the
hidden social rules which may affect such phenomenon.


Bibliography

Antelmi D. (2006),
Il discorso dei media
, Carocci, Roma.

Antinucci F. (2007),
Musei virtuali. Come non fare innovazi
one tecnologica
, Laterza, Roma.

Boyd

D.

(2007), “Why Youth (Heart) Social Network Sites: The Role of Networked Publics.” in D.
Buckingham (eds.):
Teenage Social Life. Youth, Identity and Digital Media
, MIT Press, Cambridge, MA,
pp. 119

142.

Campbell H. A.

(2010), “
Religious Authority and the Blogosphere”, in
Journal of Computer
-
Mediated
Communication

15
(2010) 251

276

Capaldi D., Ilardi E. e Ragone G. (2008),
Comunicare la memoria.
Le istituzioni culturali europee e la rete
,
Liguori, Napoli.

Carey S., Jef
frey R. (2006), “Audience Analysis in the Age of Engag
ement”, in Proceedings of
Museum and the Web 2006, Archives & Museums Informatics, Pittsburgh, Canada.

Chan S. (2007), “Radical Trust: The State of the Museum Blo
gosphere”, in Proceedings of Museum
and
the Web 2007, Archives and Museum Informatics, Pittsburgh (CA).

Colombo F. e Murru M.F. (2007), “Weblogs between Counteri
nformation and Power: an Italian
Case History”,
Observatorio (OBS*) Journal, 2 (2007), 001

013, disponibile al
http://obs.obercom.pt/index.php/obs

de Meijere A. (1996),
MAGNETS ( Museum and Art Galleries New Technologies Study)
, ed.
Videomuseums, Centre George Pompidou, Paris.

Foucault M. (1969),
L'archéologie du savoir,
Gallimard, Pari
s

(trad.it.,

L'archeologia del sapere,
Rizzoli,
1971)

Grabill J. T., Stacey P. e Wittenauer K. (2009), “Take Two: A Study of the Co

Creation of
Knowledge on Museum 2.0 Sites”, in Proceedings of Museum and the Web 2009, A
rchives and
Museum Informatics, Pitt
sburgh (CA).

Granelli A. e Traclò F. (2006),
Innovazione e cultura. Come le tecnologie digitali potenzieranno la rendita del
nostro patrimonio culturale
, Edizioni Il Sole 24 Ore, Milano.

Jenkins H. (2006),
Fans, Blogger, and Gamers. Exploring Participatory

Cultures
. New York: New York
University.
Trad. It.
Fan, Blogger e videogamers. L’emergere delle culture partecipative nell’era digitale
.
Milano:
Franco Angeli, 2008.

Liu, A., et al. (2010), “Dulwich OnView: A Museum Blog Run by the Community for the
Commu
nity”. In J. Trant and D. Bearman (eds).
Museums and the Web 2010: Proceedings
. Toronto:
Archives & Museum Informatics.

Kenix L. J. (2009), “Blogs as Alternative”, in
Journal of Computer
-
Mediated Communication
14
(2009)
790

822

Meraz S. (2009), “
Is There
an Elite Hold? Traditional Media to Social Media Agenda Setting
Influence in Blog Networks”
, in
Journal of Computer

Mediated Communication, 14

(2009), 682

707.

Micelli S., Legrenzi P. e Moretti A. (1998), “Musei virtuali, Internet e domanda di beni cultura
li”,
Sistemi Intelligenti, N. 2, ag
osto 1998, Il Mulino.

Pasquali F. e Scifo B. (2004), (a cura di),
Consumare la rete. La fruizione di Internet e la navigazione nel web
,
Vita e Pensiero, Milano.

Pelleg D., Moore A. (2000). “X
-
means: Extending K
-
means with

Efficient Estimation of the
Number of Clusters”. In proceedings of the Seventeenth International Conference on Machine
Learning (ICML), San Francisco, pages 727
-
734.

Russo A., Watkins J., Kelly L. e Chan S. (2006), “How will social media affect museum
com
munication?”, In Proceedings of Nordic Digital Excellence in Museums (NODEM), Oslo,
No
rway.

Salton G. ,Buckley C. (1988). “Term
-
weighting approaches in automatic text retrieval”. In
Information Processing and Management,

pages 513

523.

Samis P. (2008), “Wh
o has the responsibility for saying what we see? Mashing up M
useum, Artist,
and Visitor Voices, On

site and On

line” in Proceedings of Museum and the Web 2008, Archives
and Museum Informatics, Pittsburgh (CA).

Totaro A. (2008), “Dinamiche di interrelazion
e tra blogofs
era e mediasfera”, Quaderni del
C.I.R.S.D.I.G. n. 29, 2008.

Trammel K.D., Tarkowski A., Sapp A..M. (2006), “Rzeczp
ospolita blogòw [Republic of Blog]:
Examining Polish Bloggers Through Content Analysis”,
Journal of Computer

Mediated Communicat
ion
,
11(3).

Xenos M. (2008), “New Mediated Deliber
ation: Blog and Press Coverage of the Alito
Nomination”, Journal of Computer Mediated Communication 13 (2008), 485


503.