We offer a new in-depth methodology for looking at how
the use of blogging software delineates and normalizes the
blogger’s creation of posts and, by extension, the creation of
her blog self. The simple choice of software is not simple at
all and in fact has a great infl uence on the shape a blogger’s
identity will take through the interface, program design,
and data structures imposed on her by the software. This
primarily technical discussion of a topic seldom considered
in studies which look at the cultural impact of blogging, will
illuminate the inner workings of the medium and gives due
credence to Marshall McLuhan’s argument that “the ‘message’
of any medium or technology is the change of scale or pace or
pattern that it introduces into human affairs”.
Technology plays such an integral part in distinguishing the blog
as a new medium, apart from that of written (paper-based)
text; as a result, a study of the blog that does not investigate
its infrastructure is incomplete. Critics of blogging scholarship
point out the lack of technical discussion around the software
used by bloggers as a weakness (Scheidt 2005, Lawley 2004).
The criticism is valid; critics’ attention has been focused on
the output of the blogs, categorizing them as an extension of
existing genres, whether that be of the diary (Rak 167) or the
newspaper (Koman). This study serves as a response to the
criticism, and aims to start the discussion by looking into the
dark recesses of the software, databases, and code to illustrate
just how infl uential infrastructure is in defi ning identity.
Programmers do not care about the content of any given
blog. The people who develop Movable Type, Blogger,
LiveJournal, and Wordpress are developing software which
helps making blogging a much simpler process, and they do
listen to customer requests for features. But the developer
is not concerned whether your blog will be an online journal,
a political commentary, or a collection of cat pictures – what
she is concerned about is memory allocation, disk usage, and
transaction speed. Every shortcut taken in the source code,
every data type or archiving scheme not supported, every
function written, and every decision made by the programmer
to achieve these goals has an infl uence on the interface, and
therefore on the content the blogger produces. Despite
working at an indifferent distance, the developer heavily
infl uences the blog – and by extension, blogger’s identity – by
the decisions she makes when she codes the software. The way we structure language helps create meaning;
likewise, the way in which it is stored also has meaning. To
the programmer, language is nothing more than a set of bits
and data types, which must be sorted into different containers.
How the programmer deals with data affects how she creates
the interface; if she has no data structure in place to handle
a certain kind of information, she cannot request it from the
user in the interface. The data structure is created through
a process called normalization – breaking data down into its
smallest logical parts. Developers normalize data in order
to make it easier to use and reuse in a database: the title of
your blog entry goes in one container; the body text goes
into another, and so on. The structure of the data does not
necessarily match the structure of its original context, however.
Although a title and the body text are related to the same
entry, no consideration is given by the developer as to whether
one comes before the other, whether it should be displayed
in a specifi c style, or if one has hierarchical importance over
the other in the page. The data structure is dictated by the
individual pieces of data themselves. The developer takes the
data within each of these containers and stores it within a
database. This may be a simple database, such as a CSV1 or
Berkeley DB2 fi le, or it may reside within a more complex
relational database such as MySQL or Microsoft SQL Server.
Within the database exists a series of tables, and within each
table resides a series of fi elds. A table holds a single record
of data – a blog entry – and the table’s fi elds hold properties
of that data, such as the title or entry date. Figure 1 illustrates
an example of the above; a developer has created an Entries
table with the fi elds EntryID3, Title, Date, BodyText, ExtendedText,
Keywords, Category, and Post Status.
When is possible, such as with the Category and Post Status
fi elds, the developer will actually replace a string (alphanumeric)
value with a numeric pointer to the same data within another
table in the database. For example, an author may create a set
of categories for her blog (such as “Personal Life,” “School,” et
cetera, which are stored in a separate database table named
Categories and associated with a unique ID (CategoryID).
When an entry is marked with the Personal category, the
software queries the database to see what the CategoryID
of the Personal category is in the Categories table, and places
that in the Category fi eld in an entry’s record in the Entries
table (see Figure 2). This sets up a series of relations within
a database, and helps keep the database smaller; an integer
takes far less space in the database than a string: 1 byte to
store a single-digit integer, compared to 8 bytes for the string
“Personal”; when you start working with hundreds of entries,
this difference adds up quickly. It is also easier to maintain; if
you want to rename the “Personal” category to “Stories from
the woeful events of my unexaggerated life” for example, you
would only have to update the entry once in the Categories
table; because it is referenced by its CategoryID in each entry,
it will automatically be updated in all records that reference it.
By abstracting often-used data such as a category into separate
database tables, data can be reused within the database, which
in turn keeps the size of the database smaller. If we know
we will be referring to a single category in multiple entries, it
makes sense to create a table of possible categories and then
point to their unique identifi er within each individual entry.
Each fi eld within a database table is confi gured to accept a
specifi c format of information known as a data type. For
example, the Date fi eld in the Entries table above would be
given a data type of DATETIME,4 while the Category fi eld would
be given a data type of INT (to specify an integer value). The
body text of an entry would be placed in a binary data type
known as the BLOB, since this is a type of data whose size
is variable from record to record. Normalization conditions
data to its purpose, and ensures that the developer always
knows what kind of data to expect when he or she retrieves
it later. It also has the benefi t of loosely validating the data by
rejecting invalid data types. If an attempt to store a piece of
INT data in the Date fi eld is made, it will trigger an error, which
prevents the data from being misused within an application.
The decisions made by the developer at this point, which
involve confi guring the tables and fi elds within the database,
ultimately determine what will appear in the blog’s interface.
If tables and fi elds do not exist in the database to support
categorization of an entry, for example, it is unlikely to appear in
the interface since there is no facility to store the information
(and by extension, not prompt the blogger to categorize her
The interface gives the blogger certain affordances, something
Robert St. Amant defi nes as “an ecological property of the
relationship between an agent and the environment” (135).5
Amant describes affordance as a function we can see that is
intuitive: “we can often tell how to interact with an object or
environmental feature simply by looking at it, with little or no
thought involved” (135, 136) – for example, we instinctively
know not only what a chair is for, but the best way to make
use of it. St. Amant further breaks down the affordance
into four separate affordance-related concepts: relationship,
action, perception, and mental construct (136-7). He goes
on to discuss how to incorporate the idea of affordance into
developing a user interface, focusing on action and relationship.
The last of these concepts, affordance as a mental construct,
is most relevant to our discussion. St. Amant writes “these
mental affordances are the internal encodings of symbols
denoting relationships, rather than the external situations
that evoke the symbols” (137). In the authoring of the blog,
the affordance of developing identity cannot be pinned on a
single HTML control or text box; it is the process as a whole.
LiveJournal and DiaryLand, for example, have the affordance
of keeping a personal journal, or online diary. Blogger has
the affordance of developing identity in a broader way by not
necessarily focusing it on an autobiographical activity. The
interface leads the blogger into a mode of writing through
the affordances it provides The infrastructure of the blog is its
most fundamental paratextual element creating a mirror for
the blogger to peer into, but it is the blogger that makes the
decision to look. Notes
1 Comma-Separated Values. Each record of data consists of a single,
unbroken line within a text fi le, which contains a series of values
– each separated by a comma or other delimiter. An example of a
CSV fi le for our entry would look like the following: EntryID,Title,Da
te,BodyText,ExtendedText, Keywords,Category, PostStatus
1,My Entry,12/15/2006,This is the entry,,personal,Personal,Published
2,My Other Entry,12/20/2006,Look – another entry,And some
2 Berkeley DB is a fi le-based database structure, which offers some
basic relational mechanisms, but is not as robust or performant as
other database systems.
3 For compatibility with multiple database systems, spaces are
generally discouraged in both table and fi eld names.
4 For the purposes of this section, I will use MySQL Data types. Data
types may vary slightly between different database applications.
5 As an interesting historical footnote, St. Amant wrote about the
affordance in user interfaces around the time the fi rst blog software
packages were released in 1999.
Bolter, J.D. & Grusin, R. Remediation: Understanding New Media.
Cambridge, MA: The MIT Press, 1999
Lawley, Liz. “Blog research issues” Many 2 Many: A group
weblog on social software. June 24, 2004. Online. <http://many.
Accessed October 23, 2007.
Koman, Richard. “Are Blogs the New Journalism” O’Reilly
Digital Media. January 8, 2005. Online. Accessed April 8, 2007.
McLuhan, Marshall. Understanding Media: The Extension of
Man. Cambridge, Mass: MIT Press, 1994.
Rak, Julie (2005, Winter). “The Digital Queer: Weblogs and
Internet Identity.” Biography 28(1): 166-182.
Scheidt, Lois Ann. “Quals reading – Outlining in Blood”
Professional-Lurker: Comments by an Academic in
Cyberspace. Online. Accessed May 23, 2005. <http://www.
Serfaty, Viviane. The Mirror and the Veil: An Overview of
American Online Diaries and Blogs. Amsterdam & New York:
St. Amant, Robert. “Planning and User Interface Affordances”
Proceedings of the 4th international conference on Intelligent user
If this content appears in violation of your intellectual property rights, or you see errors or omissions, please reach out to Scott B. Weingart to discuss removing or amending the materials.