deletions | additions       diff --git a/SPECS-continues.tex b/SPECS-continues.tex  index 1a6f824..bd72135 100644  --- a/SPECS-continues.tex  +++ b/SPECS-continues.tex 
...
These may include randomness or simple unexpected events. The  trigger should be determined independently from the end result.}  \item[\emph{Bridge}] \emph{The system uses reasoning techniques  associated with serendipitous discovery that support a process of invention  -- e.g. abduction, e.g.~abduction,  analogy, conceptual blending -- and/or social or otherwise externally enacted  alternatives.} alternatives -- to create a bridge from the trigger to a result.}  \item[\emph{Result}] \emph{A novel result is obtained, which is  evaluated as useful, by the system and/or by an external source.}  \end{description} 
...
\begin{quote} {\em Test your serendipitous system against the standards stated in Step 2 and report the  results.}\end{quote}  We will develop several examples of the application of this framework  in Section \ref{sec:computational-serendipity}.  %%  In order to develop connections with our theoretical framework, and %%  because existing experiments have not been particularly strong, we %%  focus on a thought experiment in the %%  Section \ref{sec:computational-serendipity} detailing some of the %%  outcomes we would like to see, and some of the risks. diff --git a/background.tex b/background.tex  index 546fe39..0502ac2 100644  --- a/background.tex  +++ b/background.tex 
...
\section{Background} [Note, this looks different on Authorea than it does in normal LaTeX.] \label{sec:background}  \input{SPECS-begins.tex}  diff --git a/bibliography/biblio.bib b/bibliography/biblio.bib  index 1f6271e..127ea87 100644  --- a/bibliography/biblio.bib  +++ b/bibliography/biblio.bib 
...
@misc{poetry-workshop,  title={{C}omputational {P}oetry {W}orkshop: {M}aking {S}ense of {W}ork in {P}rogress},  author={J. Corneli and A. Jordanous and R. Shepperd and M. T. Llano and J. Misztal and S. Colton and C. Guckelsberger},  url={http://metameso.org/~joe/docs/poetryICCC-wip.pdf},  year={2015}  }  @article{engineering-serendipity,  title={Engineering {S}erendipity},  author={Lindsay, Greg},  journal={New York Times},  day={7},  month={April},  year={2013},  url={http://www.nytimes.com/2013/04/07/opinion/sunday/engineering-serendipity.html}}  @article{creativity-crisis,  title={The {C}reativity {C}risis},  author={Bronson, Po and Merryman, Ashley}, 
...
@book{gabriel2002writer,  Author = {Gabriel, Richard P},  Publisher = {Addison-Wesley Longman Publishing Co., Inc.},  Title = {{W}riter's {W}orkshops and the {W}ork of {M}aking {T}hings}, {T}hings: Poetry, Patterns{\ldots}},  Year = {2002}}  @article{rowe1994creativity,  diff --git a/conclusion.tex b/conclusion.tex  index 39f2efb..0c42a12 100644  --- a/conclusion.tex  +++ b/conclusion.tex 
...
\section{Conclusion} \label{sec:conclusion}  In his treatise on logic and scientific method, W. Stanley Jevons  wrote:  diff --git a/introduction.tex b/introduction.tex  index 56061c9..f0472a1 100644  --- a/introduction.tex  +++ b/introduction.tex 
...
Although computational creativity is well studied in both theory and  practice, the role of \emph{serendipity} has often not been discussed  in this field -- even though serendipity hasoften  played a well-documented  role in historical instances of scientific and technical creativity. One reason for this omission  may be that the field of computational creativity has tended to focus on artistic creativity, while neglecting other fields. creativity.  But serendipity is also increasingly seen as  relevant to within  the arts \cite{mckay-serendipity} and across human other  creative endeavour \cite{kakko2009homo}. enterprises  \cite{kakko2009homo,engineering-serendipity}: it is managed and  encouraged with methods ranging from architecture to data science.  %  An interdisciplinary perspective on the phenomenon of serendipity  promises shared further  illumination. Here, we consider its the  potential for  formalising this concept and investigate its utility  as a new framework for computational creativity. This is appropriate, since many instances of serendipity centre Serendipity centres  on reevaluation. reassessment.  For example, a non-sticky ``superglue'' that no one was quite sure how to use turned out to be just the right ingredient for 3M's Post-it\texttrademark\ notes. %  Serendipity is related, firstly, to deviations from expected or  familiar patterns, and secondly, to new insight. 
...
anymore}.'' \cite{van1994anatomy}  \end{quote}  We believe that serendipity is not so mystical as such statements  might seem to  imply, and in Section \ref{sec:computational-serendipity} we will show how it is possible to reinterpret van Andel's ``patterns of serendipity'' in computational settings.The real problem with computers is not that they only do what they're  told, but that the act of programming forces us to confront the  emergence of the new \cite{mead1932philosophy}.  %  Indeterminacy forms an important part of any proposal for  ``intelligent machines'', after Turing:  \begin{quote}  ``\emph{They will make mistakes at times, and at times they may make  new and very interesting statements, and on the whole the output  of them will be worth attention to the same sort of extent as the  output of a human mind}.'' \cite{turing-intelligent}  \end{quote}  Serendipity has played a role in the large-scale history of the  computing field \cite{de2013turing} and in artistic applications of  computer technology \cite{reichardt1969cybernetic}. We aim to clarify  the role it has to play in the future development of computational  creativity.  First, in Section \ref{sec:background} we present our formal  definition of serendipity, and examine related work that has applied  the concept of serendipity in a computational context. Then in  Section \ref{sec:literature-review}, we survey the broad literature on  serendipity, making connections from historical examples of  serendipitous discovery and invention to our formal model. Section  \ref{sec:computational-serendipity} then presents case studies and  thought experiments in terms of this model. Section  \ref{sec:discussion} offers recommendations for researcher working in  the computational creativity, and describes our own plans for future  work. Section \ref{sec:conclusion} reviews the argument and  summarises the limitations of our analysis.  diff --git a/literature.tex b/literature.tex  index 42d0e60..864c091 100644  --- a/literature.tex  +++ b/literature.tex 
...
particular, we point out key conditions of serendipity, their  components and general characteristics, including environmental  factors. The structure of this section follows and updates an earlier  survey from \citeA{pease2013discussion}. \citeA{pease2013discussion}, drawing connections with our  formal model described above.  \subsection{Etymology and selected definitions} \label{sec:overview-serendipity}  The English term ``serendipity'' derives from the 1302 long poem ``Eight Paradises'', \emph{Eight Paradises},  written in Persian by the Sufi poet Am\={\i}r Khusrow in Uttar Pradesh.\footnote{\url{http://en.wikipedia.org/wiki/Hasht-Bihisht}} In the English-speaking world, its first chapter became known as ``The Three Princes of Serendip'', where ``Serendip'' represents the Old Tamil-Malayalam word for Sri Lanka (%{\tam சேரன்தீவு}, \emph{Cerantivu}), ``island of the Ceran kings.''  The term ``serendipity'' is first found in a 1557 1757  letter by Horace Walpole to Horace Mann: \begin{quote}  \emph{``This discovery is almost of that kind which I call serendipity, a very expressive  word} \ldots \emph{You will understand it better by the derivation than by the 
...
%  Classic definitions require the investigator not to be aware of the problem they serendipitously solve, but this criterion has largely dropped from dictionary definitions. Only 5 of Roberts' collected definitions explicitly say ``not sought for.'' Roberts characterises ``sought findings'' in which an accident leads to a discovery with the term \emph{pseudoserendipity} \cite{chumaceiro1995serendipity}.  %  While Walpole initially described serendipity as an event (a discovery), event,  it has since been reconceptualised as a psychological attribute, a matter of sagacity on the part of the discoverer: a ``gift'' or ``faculty'' more than a ``state of mind.'' Only one of the collected definitions, from 1952, defined it solely as an event, while five define it as both event and attribute. However, there are numerous examples that exhibit features of  serendipity which develop on a social scale rather than an individual 
...
It is important to note that serendipity is usually discussed within  the context of \emph{discovery}, rather than \emph{creativity},  although in typical parlance these terms are closely related  \cite{jordanous12jims}. In our definition of serendipity, we have  made use of  Henri Bergson's distinction will be useful in  what follows: distinction:  \begin{quote}  ``\emph{Discovery, or uncovering, has to do with what already exists,  actually or virtually; it was therefore certain to happen sooner  or later. Invention gives being to what did not exist; it might  never have happened.}''~\cite{bergson2010creative}  \end{quote}  Serendipity, as As  we understand the term, have indicated serendipity  would seem to require features of both; that is, the discovery of something unexpected and the invention of an application for the same. We must complement analysis \emph{analysis}  with synthesis \emph{synthesis}  \cite{delanda1993virtual}. The balance between these two features will differ from case to case.In the following section,  we will elaborate on the characteristics of serendipity with  particular reference to classic examples.  The story In the next subsection we will review several historical examples.  First, one further point should be made with reference to the ``The  Three Princes  of ``Eight Paradises'' was also Serendip''. Prior to Walpole's coinage, this story  had been  adapted by Voltaire  into an early chapter of Voltaire's Zadig, \emph{Zadig},  and in turn ``the method of Zadig'' informed subsequent approaches both to fiction writing and natural science. This method is, to be sure, is  rooted firstly  in discovery: \begin{quote}  ``[H]\emph{e pry’d ``[Zadig] \emph{pry’d  into the Nature and Properties of Animals and Plants, and soon, by his strict and repeated Enquiries, he was  capable of discerning a Thousand Variations in visible Objects,  that others, less curious, imagin’d were all  alike.}''~\cite[pp. 21--22]{zadig}  \end{quote}  \noindent However the essential thing is that Secondly,  fromthese various  disparate observations, Zadig is often  able to assemble a coherent picture: \begin{quote}  \emph{It was his peculiar Talent to render Truth as obvious as  possible: Whereas most Men study to render it intricate and  obscure.}~\cite[p. 54]{zadig}  \end{quote}  Similarly, but in reverse, a coherent picture may be reduced to  fragmented pieces each of which may  tell a very  different story from the whole. This is illustrated in Zadig's misadventure with a broken  tablet, in which one fragment of a poem of praise reads as treasonous  provocation.  In enumerating describing  the various features of serendipity below, we will draw connections with the schematic diagram presented in Section \ref{specs-overview}, in order to best present unfold  the multifacetedbut  coherent  notion of serendipity. \subsection{Connections between prior literature on serendipity and our formal definition} \label{sec:connections-to-formal-definition}  Each of the features described below, using an example drawn from the  literature on serendipity, with connections to one part of our diagram.  \subsubsection*{Key condition for serendipity}  \begin{itemize} 
...
write that: ``To be creative requires divergent thinking (generating  many unique ideas) and then convergent thinking (combining those  ideas into the best result).'' Accordingly $T^\star$ may be thought  of as an evolving vector of interesting possibilities. possibilities or ``strategic data'' \cite[p. 507]{merton1948bearing}.  In de Mestral's case, the initial idea of a hook-and-loop fastener  occurred in 1941, 1941 --  followed by a full  decade of experimentation before he was ready to file a patent claim.} claim. }  \end{itemize}  \subsubsection*{Components of serendipity}  diff --git a/recommendations.tex b/recommendations.tex  index 4e7fe83..a0a2c78 100644  --- a/recommendations.tex  +++ b/recommendations.tex 
...
\subsection{Recommendations} \label{sec:recommendations}  In the diagrammatic formalism advanced in  \cite{colton-assessingprogress}, we spoke about progress with  \emph{systems} rather than with \emph{problems}. It would be a useful  generalisation of the formalism -- and not just a simple relabelling  -- to tackle problems as well.  %  Figueiredo and Campos \citeyear{Figueiredo2001}, for example, describe  serendipitous ``moves'' from one problem to another.  %  However, progress with problems does not always mean transforming a  problem that cannot be solved into one that can. Progress may also  diff --git a/related-work.tex b/related-work.tex  index d78ee33..d52e606 100644  --- a/related-work.tex  +++ b/related-work.tex 
...
\subsection{Related work} \label{sec:related}  \textbf{[Since we are now have the formal definition, let's be sure  that we're sufficiently thorough in connecting this related work  back to it -- and pointing out what still needs to be done.]}  Paul Andr{\'e} et al.~\citeyear{andre2009discovery} look at serendipity from a design point of view. They These authors also  propose a two-part model, in which what we might call chance+curiosity have called \emph{discovery} above  exposes the unexpected, and  sagacity+value is determined by another subsystem. This corresponds  to Bergson's distinction between \emph{discovery} and while  \emph{invention}(see Section \ref{sec:overview-serendipity}). One survey related to  the first phase  is \cite{foster2003serendipity}. the responsibility  another subsystem that finds applications.  According to Andr\'e et al., the first phase is the one that has most frequently been automated, but they suggest that computational systems should be  developed that support both aspects. Their specific suggestions focus  on representational features: \emph{domain expertise} and a  \emph{common language model}. We've advocated Although tremendously useful when they are available, these features  are not always enough to account  for serendipitious events. Using the  terminology we introduced above, these features seem to exemplify  aspects of the \emph{prepared mind}. However, as we mentioned above,  the \emph{bridge} is  a more  experimentally-based approach distinct process  that does mental preparation can  support, but  not directly rely on shared  understandings. always fully determine.  For example, participants in a Writers Workshopin  poetry  may not ``understand'' a possess a very limited understanding of each  other's aims or of the work they are critiquing, and may as a  consequence talk past  one another but can still find to a greater or lesser degree --  while nevertheless finding  the experience overall process  of participating in the workshop rewarding. itself illuminating and rewarding (often precisely because  such misunderstandings elucidate poor communication choices!).  Various social strategies, ranging from Writers Workshops to open  source software, pair programming, and design charettes  \cite[p. 11]{gabriel2002writer} have been developed to exploit similar  emergent effects to develop new insights, and to develop \emph{new}  shared language. In \cite{poetry-workshop}, we investigate the  feasibility of using designs of this sort in multi-agent systems that  learn by sharing and discussing partial understandings. This earlier  paper remains broadly indicative, however, and the ideas it describes  can see considerable benefit from the more formal thinking we develop  in the current work.  The issue of designing for serendipity has also been taken up recently  by Deborah Maxwell et al.~\cite{maxwell2012designing}, al.~\citeyear{maxwell2012designing},  in their description of a prototype of the {\sf SerenA} system. This system is  designed to support serendipitous discovery for its \emph{users} (human) users  \cite{forth2013serena}. The authors rely on a process-based model of  serendipity \cite{Makri2012,Makri2012a} that is derived from user  studies, including interviews with 28 researchers, looking for  instances of serendipity from both their personal and professional  lives. This material was coded along three dimensions:  \emph{unexpectedness}, \emph{insightfulness}, and \emph{value}. This  work research  aims to support the process of forming bridging connections that  eventuate in an from unexpected encounter to a previously  unanticipated but  valuable outcome. They particularly focus on the acts of \emph{reflection} that foment both the creation of a  bridge and estimates of the potential value of the result.Both pattern-building  activities and the practice of fomenting thought by structured  encounters in Writers Workshops can be understood to contribute to the  theory and practise of reflection\footnote{As with creativity and  serendipity, in order to carry out concrete evaluations of automated  reflection we may well ask ``what, exactly, are we looking for as  evidence of reflection?'' \cite{rodgers2002defining}. A detailed  answer derived from the classic work of John Dewey  \citeyear{dewey1997we} is explored in \cite{rodgers2002defining}.}  Although this touches on all of the features of our model,  {\sf SerenA} is a system like nevertheless matches  the ones described description offered  by Andr{\'e} et al.~\cite{andre2009discovery}, in which al.~\citeyear{andre2009discovery} of discovery-focused systems:  the user isexpected to have  the primary agent with a prepared mind. Accordingly it is the  user that undergoes an  ``aha'' moment, moment  and take takes  the creative steps. The steps to  realise the result; the  computer is mainly used to facilitate this; and as indicated above this this.  The primary computational method  is usually  done by searching to search  outside of the normal search parameters in order  to engineer potentially serendipitous (or at least pseudo-serendipitous) encounters. Another earlier related  example of this sort of system is {\sf Max}, created by Figueiredo and Campos \citeyear{Campos2002}. The user emailed {\sf Max} with a list of interests and {\sf Max} would find a webpage that may be of interest to the user. Other search-related  examples Similar systems with  support for serendipitous  discovery involve  searching for analogies (\cite{Donoghue2002} and  \cite{Donoghue2012}) \cite{Donoghue2002,Donoghue2012})  and content \cite{Iaquinta2008}. In earlier joint work \cite{colton-assessingprogress}, we presented a  diagrammatic formalism for evaluating progress in computational  creativity. It is useful to ask what serendipity would add to this  formalism, and how the result compares with other attempts to  formalise serendipity, notably Figueiredo and Campos's  \citeyear{Figueiredo2001} `Serendipity Equations'.%  In \cite{stakeholder-groups-bookchapter}, this work,  Figueiredo and Campos describe serendipitous ``moves'' from one  problem to another, which transform a problem that cannot be solved  into one that can. In our diagrammatic formalism,  we advanced several  hypotheses related spoke about  progress with \emph{systems} rather than with \emph{problems}. It  would be a useful generalisation of the formalism -- and not just a  simple relabelling -- for it to be able  to tackle problems as well.  However, progress with problems does not always mean transforming a  problem that cannot be solved into one that can. Progress may also  apply to growth in  the development ability to \emph{posit} problems. In keeping  track  of progress, it would be useful for system designers to record  (or get their systems to record) what problem a given system solves,  and  the computational creativity  field. Again, we should ask here how degree to which the computer was responsible for coming up  with this problem. The relationship between  serendipity contributes. We  discuss these points in and novel  problems receives considerable attention here, since we want to  increasingly turn over responsibility for creating and maintaining a  prepared mind to  the following section. machine.  diff --git a/serendipity-in-computational-context.tex b/serendipity-in-computational-context.tex  index a032faf..e5005c3 100644  --- a/serendipity-in-computational-context.tex  +++ b/serendipity-in-computational-context.tex 
...
% Why is it appropriate (formal spec e.g. considering externalities)  % what is the logic of the strategy by which it can be carried out.  \textbf{[We will have to make good on our promise to ``reinterpret van  Andel's `patterns of serendipity' in computational settings'' or  else drop it from the introduction.]}  % \newpage  \subsubsection*{Key condition for serendipity} 
...
\emph{might} operate in a serendipitous fashion, as well as what  limitations it runs into in the process.]}  \subsection{On evaluating a \subsection{A  Writers Workshop for Systems} \textbf{[It would be good to go back over our other paper and make  sure we make good on the idea in the Related Work section of the  current paper that ``This earlier paper remains broadly  indicative, however, and the ideas it describes can see  considerable benefit from the more formal thinking we develop in  the current work.'']}  \begin{itemize}  \item \textbf{In particular: at least one of the reviewers found the  Writers Workshop ``technologically unrealistic'' or similar, so  let's try to make sure we're not overpromising. I think the other  paper makes it all fairly realistic.}  \end{itemize}  \paragraph{Writers Workshop: Prepared mind.}  Each contributing system should come to the workshop with at least a  diff --git a/serendipity.tex b/serendipity.tex  index 90ba573..4691209 100644  --- a/serendipity.tex  +++ b/serendipity.tex 
...
\tableofcontents  \newpage  \input{introduction.tex}  \input{background.tex}  \input{related-work.tex}