Ahnert et al 2020: Difference between revisions

From Whiki
Jump to navigation Jump to search
(Created page with "https://www.cambridge.org/core/elements/network-turn/CC38F2EA9F51A6D1AFCB7E005218BBE5")
 
No edit summary
 
Line 1: Line 1:
https://www.cambridge.org/core/elements/network-turn/CC38F2EA9F51A6D1AFCB7E005218BBE5
https://www.cambridge.org/core/elements/network-turn/CC38F2EA9F51A6D1AFCB7E005218BBE5
== The Network Turn - Introduction ==
Albert-László Barabási and Réka Albert published a scientific article entitled ‘Emergence of Scaling in Random Networks’ (Barabási & Albert, 1999). It argued that a wide variety of seemingly heterogeneous networks, such as power grids, social networks, and the World Wide Web, exhibit nearly identical distributions of connectivity, and it offered an elegant model that explained how these distributions might arise…
1
Lombardi and Barabási’s work is part of what we call the ‘network turn’. This turn cannot be attributed to either the artist or the scientist; they are but two examples of a whole host of converging thoughts and practices around the turn of the new millennium – the zeitgeist of the networked age.
3
This book not only argues that arts and humanities scholars can use the same kind of visual and quantitative analysis of networks to shed light on the study of culture; it also contends that the critical skills native to humanistic inquiry are vital to the theorisation and critique of our networked world.
4
We con- tend that networks are a category of study that cuts across traditional academic boundaries and that has the potential to unite diverse disciplines through a shared understanding of complexity in our world – whether that complexity pertains to the nature of the interactions of proteins in gene-regulatory net- works or to the network of textual variants that can reveal the lineage of a poem.
4
The artist and scientists use connectivity to make sense out of data: a representation of knowledge that relies on abstraction. Both produce results that are seductive in their elegance and simplicity. Networks are by definition an abstraction into a system of nodes and edges. Nodes are entities; edges are the relationships between them.
5
This book does not call for arts and humanities scholars to accept unquestioningly frameworks and methods developed in the field of network science. Rather, it argues that the discourse and analysis of networks can move forward only through collaboration and exchange at the interface of computational method, humanistic inquiry, and design practice.
6
Although researchers with standard huma- nities training will likely need to acquire some new skills to engage with the computational challenges of network visualisation and quantitative analysis, we contend that they already have a set of skills that are key to the develop- ment of the interdisciplinary practice of network analysis. This is not just about receiving wholesale methods and theories developed in the computational and social sciences; rather, the critical skills developed in the arts and humanities are needed to complicate and nuance the current ways in which data are collected, modelled, and queried in the field of network science.
7
== The Network Turn - Part I Frameworks - 1 Networks Are Always Metaphorical ==
In network science, researchers utilise networks as a formalised abstraction that permits computational analysis. In the humanities, by comparison, scholars largely employ networks as a metaphor.
13
Network science as a field takes the abstract network as a starting point; the process of abstraction often belongs to another domain, namely that in which the network data originates.
13
This chapter argues that researchers who employ networks as a meta- phor (traditionally those in the arts and humanities) ought to be familiar with the mathematical formalisations. Conversely, scholars wedded to the computational power of quantitative network analysis should be aware that its power derives from its reliance on the metaphorical dimension and an act of interpretation.
14
The movement into the digital realm, however, does something to the way we think about networks. It seems to make scientists custodians of the knowledge we have about networks, even though the systems they are analysing are historically the intellectual domains of very different disciplines. By thinking of them as something that can be measured mathematically, they no longer seem metaphorical but real and knowable. However, science is not the saviour of these other disciplines; rather, these discoveries depend on the convergence of numerous disciplines that have zeroed in on one way of understanding the world.
21
Before information and commu- nications technology made it possible to gather and analyse large-scale net- work data in the 1990s, such networks were often assumed to be simple random networks in which links exist with uniform probability. Barabási and Albert discovered that many real-world networks have a very different connectivity – one in which a small number of nodes are very highly connected, a larger number of nodes are reasonably well connected, and the vast majority are poorly connected.
22
The metaphorical dimension of the network allows network scientists to imagine possible ways of navigating mathematical spaces that are both conceptually and topologically vast. The lens of metaphor, therefore, could be described as limiting, but this limitation is productive, giving researchers somewhere to begin their explorations.
23
== The Network Turn - Part I Frameworks - 2 Historical Threads ==
Humanities scholars may seek network thinking for its ability to cut through hierarchies, allowing us to draw threads between and through the geopolitical hegemonies that are often reflected in the construction of physical archives. Networks might give us the language to speak truth to power in configurations learned from and reminiscent of the US civil rights movement, the Iranian Green Movement, or the Extinction Rebellion.
25
The network turn has become complicit in various hegemonic power structures. As centres of power take networks and network theory increasingly seriously, they build ever more network assumptions into their systems.
25
We engage with networks not because it would be irresponsible not to, true though that might be, but because it becomes increasingly impossible not to.
26
…when we collectively settle on particular visual metaphors to order our world – like trees of knowledge – that representation radiates outward, shaping our thoughts in unexpected ways. Conversely and recursively, our understanding of our world shapes our visual metaphors…
29
As different as they were, the diagrams of eugenicists and those of H. G. Wells shared a common interpretation: that the universe has no innate order and it is left for us to create one. It was in this context that the visual lexicon
36
of the what we now call networks took hold, perhaps because they are the closest ontological metaphor to trees which require neither hierarchy nor root. The way we conceptualise and chart systems of knowledge in the twenty-first century was radically shaped by systems of notation Jacob Moreno and Helen Hall Jennings developed in the 1930s.
37
1), Moreno and Jennings devel- oped a psychological technique involving the construction and analysis of sociograms, diagrams, like the one shown in in Figure 4, that revealed complex social relations between small communities.
37
Just as the tree brought with it implicit notions of hierarchy and unity, this visual language has far-reaching interpretative implications. Firstly, such visualised networks reject hierarchies. Even when node-and- link diagrams represent hierarchical networks, those hierarchies are difficult to notice.
38
Secondly, networks embrace connectedness: in a consilient world, where knowledge collectively acts as a foundation for the whole, or in actor-network theory, where agency can be widely dispersed, this representation fits like a glove. Thirdly, networks separate ontology from essence.
38
Knowledge’s order is always left uncertain, which feeds into the fourth point: that networks have the capacity to relate situated perspectives.
38
The democratising effect of the network view of the world is perhaps most radically realised in actor-network theory (ANT).
38
The process of levelling allows us to challenge other narratives, not only social, but historical, literary, aesthetic, and linguistic, overturning assumptions about causality, hierarchy, the distribution of power, and the direction and quantity of influence.
39
In these contexts the network perspective is inherently political. There is something decidedly democratic about the initial process that disarms agency and power.
39
However, the network turn’s ability to break existing hierarchical power structures does not imply that it is an acid in which all power melts. Networks reify power along different lines, such as centrality, as evidenced by the enormous power of popular social media presences to gatekeep or to spread certain ideas.
40
As governments, tech companies, and other centres of power began to take seriously the theories and affordances of networks, they started building the assumptions of those affordances into their systems as a means to re-exert control. As systems start operating on these principles, the world contorts around them to oblige.
40
== The Network Turn - Part II Cultural Networks - 3 Culture Is Data ==
Nothing is naturally a network; rather, networks are an abstraction into which we squeeze the world. Nevertheless, almost anything can be turned into a network, whether it be the interactions between characters in Shakespeare’s plays, the dissemination of …
43
In such a context people go from being nodes to being the connecting edges between places, as in the article ‘A Network Framework of Cultural History’ (and the accompanying video), which provides an overview of the movements of humanity over the past 2,600…
46
Without the metaphorical and analytic power of networks, it is difficult to understand how the interactions of characters in Hamlet are in any way comparable to the dissemination of memes on Facebook, or the chemical compounds shared between food ingredients. What makes them compar- able is an intellectual and methodological shift by which we abstract our objects of study into data points that can be entered into a database or spreadsheet. This does not imply some shared property intrinsic to each of the subjects under study, rather it implies the widespread utility of networks as a lens through which to view many aspects of our shared world.
51
It is important to remember, however, that almost any network data set, even in the sciences, is incomplete in some sense.
52
While this of course will affect the results of any quantitative analysis, this incompleteness is less of a problem than it may seem at first, for three main reasons.
52
The first reason is that, particularly in the context of the humanities, the bias in the source data is often itself of interest to the scholar.
52
The second reason is that, even in an incomplete network, the relative importance of nodes and edges according to a given network measurement can still yield meaningful results.
53
The third and most important reason why the incompleteness of net- works does not pose an insurmountable problem is that the results of the quantitative analysis do not serve as a final result, but as a starting point for further, more detailed inquiry in the vein of traditional scholarship in the arts and humanities.
53
== The Network Turn - Part II Cultural Networks - 4 Visual Networks ==
Even with this simple overview, it quickly becomes apparent how much the arrangement of nodes influences our reading of a network.
63
What is useful about Bertin’s work is that it gives us a point of reference for reading the rhetoric encoded in design decisions.
63
It is not only that computer graphics look precise, computer graphics software is encoded with assumptions about graphical meaning rooted in Bertin’s theory of graphics as monosemic and unambiguous.
64
But we do need to be aware of the assumptions encoded in the tools we use so that we can bend them to our own needs.
64
The force-directed layouts are a class of algorithms used frequently in Gephi and the visual form with which readers will be most familiar because of their increasing ubiquity through high-volume use of Gephi and similar tools. In a force-directed layout, each element of the network is modelled as though guided by unseen forces, much as in the physical world: gravity, electromagnetic repulsion, material elasticity. Nodes may be thought of as electrons on a two-dimensional plane, forcing each other apart as they approach one another, and the edges may be conceived of as springs physically anchored between two nodes, drawing them together even as they repulsively push away.
66
Tommaso Venturini, Mathieu Jacomy, and Pablo Jensen have begun to address this challenge by explicitly defining the strengths and weaknesses of what they have called visual network analysis (VNA) and providing guide- lines for how to make effective use of algorithmic layouts to expose the topological structure of large networks…
67
The argument underlying VNA is that understanding the conceptual and mathematical underpinnings of the layout algorithms and choosing appropriate settings can result in effective, legible graph spatialisation.
67
Though data analysis is widely considered synonymous with particular forms of visualisation these days, charts and graphs were not widely accepted until they were formalised in the mid- twentieth century when computing and, specifically, computational informa- tion processing was on the rise.
68
…visualisation is not better than narrative argument or mathema- tical equations for communicating ideas, but that it provides an additional means of producing, exploring, and analysing information that has proven value in both the liberal arts and the sciences…
70
Visualisations of networks are unsettling because the graphical language used to produce them is not, in fact, precise.
70
== The Network Turn - Part III Manoeuvres - 5 Quantifying Culture ==
Numbers have the ability to capture certain attributes that cannot be gleaned simply by reading text or looking at images. Statistics can make an argument that cannot be expressed by words alone. Despite this, the quantitative is perceived as at odds with the normal practices and tropes of cultural commen- tary. Quantitative network analysis presents a challenge to traditional scholar- ship in the humanities not only in terms of methodology, but also in the ways in which we write about the findings from such a process.
74
…the use of any of these tools requires a prior mental manoeuvre of translating cultural artefacts into an abstracted form to see whether they are compatible with the input requirements of the available tools. This process of abstraction is not only a way of thinking, it also gives us algorithmic power.
75
Moreover, and as we argue in more detail in the following chapter, the research process will involve shifting back and forth across this continuum when addressing or honing a particular question.
79
…single off-the-shelf algorithms, such as the centrality measures mentioned earlier, can provide useful first insights into the data, but the likelihood of any single measure being a reasonable proxy for the specific cultural attributes or phenomena one is looking for is minimal. To find the nodes or individuals of particular interest, or to develop more nuanced research questions beyond structural network questions, the measures typically need to be combined with other quantitative or qualitative approaches.
83
…by looking at how the different categories of nodes interacted, they observed that the most prolific leaders frequently and repeatedly wrote to network sustainers, meaning that the shortest paths across the network were also the ones most frequently traversed by letters and, by implication, carriers.
84
Categories of people such as those defined for the Protestant network show how network analysis enables us to manoeuvre between different registers of quantitative analysis, from micro, to meso, to macro. Both the use of centrality measures and the construction of network ‘profiles’ are about generating a quantitative description of certain attributes of a single node as a function of their position in the entire network.
84
Centrality measures can also be used to think about the mesoscopic properties of complex networks.
84
In light of the changing ways in which culture is studied, we contend that a pressing duty is placed on the university to prepare future generations of academics by offering suitable combinations of courses in humanities subjects, programming, and statis- tical methods.
86
…how we can produce scholarship that is accessible to audiences from all the contributing disciplines. In constructing practices of research that lie at the intersection of several disciplines, we need to make sure that we have created something greater than the sum of its parts, rather than something lesser, such as an inter- disciplinary silo in which we are only speaking to other displaced scholars…
88
== The Network Turn - Part III Manoeuvres - 6 Networking the ‘Divided Kingdom’ ==
The correlative to the scales of investigation – the satellite view, the aerial view, and the archaeological dig – are the different kinds of argument that can be proffered.
95

Latest revision as of 02:37, 12 November 2022

https://www.cambridge.org/core/elements/network-turn/CC38F2EA9F51A6D1AFCB7E005218BBE5

The Network Turn - Introduction

Albert-László Barabási and Réka Albert published a scientific article entitled ‘Emergence of Scaling in Random Networks’ (Barabási & Albert, 1999). It argued that a wide variety of seemingly heterogeneous networks, such as power grids, social networks, and the World Wide Web, exhibit nearly identical distributions of connectivity, and it offered an elegant model that explained how these distributions might arise…

1

Lombardi and Barabási’s work is part of what we call the ‘network turn’. This turn cannot be attributed to either the artist or the scientist; they are but two examples of a whole host of converging thoughts and practices around the turn of the new millennium – the zeitgeist of the networked age.

3

This book not only argues that arts and humanities scholars can use the same kind of visual and quantitative analysis of networks to shed light on the study of culture; it also contends that the critical skills native to humanistic inquiry are vital to the theorisation and critique of our networked world.

4

We con- tend that networks are a category of study that cuts across traditional academic boundaries and that has the potential to unite diverse disciplines through a shared understanding of complexity in our world – whether that complexity pertains to the nature of the interactions of proteins in gene-regulatory net- works or to the network of textual variants that can reveal the lineage of a poem.

4

The artist and scientists use connectivity to make sense out of data: a representation of knowledge that relies on abstraction. Both produce results that are seductive in their elegance and simplicity. Networks are by definition an abstraction into a system of nodes and edges. Nodes are entities; edges are the relationships between them.

5

This book does not call for arts and humanities scholars to accept unquestioningly frameworks and methods developed in the field of network science. Rather, it argues that the discourse and analysis of networks can move forward only through collaboration and exchange at the interface of computational method, humanistic inquiry, and design practice.

6

Although researchers with standard huma- nities training will likely need to acquire some new skills to engage with the computational challenges of network visualisation and quantitative analysis, we contend that they already have a set of skills that are key to the develop- ment of the interdisciplinary practice of network analysis. This is not just about receiving wholesale methods and theories developed in the computational and social sciences; rather, the critical skills developed in the arts and humanities are needed to complicate and nuance the current ways in which data are collected, modelled, and queried in the field of network science.

7

The Network Turn - Part I Frameworks - 1 Networks Are Always Metaphorical

In network science, researchers utilise networks as a formalised abstraction that permits computational analysis. In the humanities, by comparison, scholars largely employ networks as a metaphor.

13

Network science as a field takes the abstract network as a starting point; the process of abstraction often belongs to another domain, namely that in which the network data originates.

13

This chapter argues that researchers who employ networks as a meta- phor (traditionally those in the arts and humanities) ought to be familiar with the mathematical formalisations. Conversely, scholars wedded to the computational power of quantitative network analysis should be aware that its power derives from its reliance on the metaphorical dimension and an act of interpretation.

14

The movement into the digital realm, however, does something to the way we think about networks. It seems to make scientists custodians of the knowledge we have about networks, even though the systems they are analysing are historically the intellectual domains of very different disciplines. By thinking of them as something that can be measured mathematically, they no longer seem metaphorical but real and knowable. However, science is not the saviour of these other disciplines; rather, these discoveries depend on the convergence of numerous disciplines that have zeroed in on one way of understanding the world.

21

Before information and commu- nications technology made it possible to gather and analyse large-scale net- work data in the 1990s, such networks were often assumed to be simple random networks in which links exist with uniform probability. Barabási and Albert discovered that many real-world networks have a very different connectivity – one in which a small number of nodes are very highly connected, a larger number of nodes are reasonably well connected, and the vast majority are poorly connected.

22

The metaphorical dimension of the network allows network scientists to imagine possible ways of navigating mathematical spaces that are both conceptually and topologically vast. The lens of metaphor, therefore, could be described as limiting, but this limitation is productive, giving researchers somewhere to begin their explorations.

23

The Network Turn - Part I Frameworks - 2 Historical Threads

Humanities scholars may seek network thinking for its ability to cut through hierarchies, allowing us to draw threads between and through the geopolitical hegemonies that are often reflected in the construction of physical archives. Networks might give us the language to speak truth to power in configurations learned from and reminiscent of the US civil rights movement, the Iranian Green Movement, or the Extinction Rebellion.

25

The network turn has become complicit in various hegemonic power structures. As centres of power take networks and network theory increasingly seriously, they build ever more network assumptions into their systems.

25

We engage with networks not because it would be irresponsible not to, true though that might be, but because it becomes increasingly impossible not to.

26

…when we collectively settle on particular visual metaphors to order our world – like trees of knowledge – that representation radiates outward, shaping our thoughts in unexpected ways. Conversely and recursively, our understanding of our world shapes our visual metaphors…

29

As different as they were, the diagrams of eugenicists and those of H. G. Wells shared a common interpretation: that the universe has no innate order and it is left for us to create one. It was in this context that the visual lexicon

36

of the what we now call networks took hold, perhaps because they are the closest ontological metaphor to trees which require neither hierarchy nor root. The way we conceptualise and chart systems of knowledge in the twenty-first century was radically shaped by systems of notation Jacob Moreno and Helen Hall Jennings developed in the 1930s.

37

1), Moreno and Jennings devel- oped a psychological technique involving the construction and analysis of sociograms, diagrams, like the one shown in in Figure 4, that revealed complex social relations between small communities.

37

Just as the tree brought with it implicit notions of hierarchy and unity, this visual language has far-reaching interpretative implications. Firstly, such visualised networks reject hierarchies. Even when node-and- link diagrams represent hierarchical networks, those hierarchies are difficult to notice.

38

Secondly, networks embrace connectedness: in a consilient world, where knowledge collectively acts as a foundation for the whole, or in actor-network theory, where agency can be widely dispersed, this representation fits like a glove. Thirdly, networks separate ontology from essence.

38

Knowledge’s order is always left uncertain, which feeds into the fourth point: that networks have the capacity to relate situated perspectives.

38

The democratising effect of the network view of the world is perhaps most radically realised in actor-network theory (ANT).

38

The process of levelling allows us to challenge other narratives, not only social, but historical, literary, aesthetic, and linguistic, overturning assumptions about causality, hierarchy, the distribution of power, and the direction and quantity of influence.

39

In these contexts the network perspective is inherently political. There is something decidedly democratic about the initial process that disarms agency and power.

39

However, the network turn’s ability to break existing hierarchical power structures does not imply that it is an acid in which all power melts. Networks reify power along different lines, such as centrality, as evidenced by the enormous power of popular social media presences to gatekeep or to spread certain ideas.

40

As governments, tech companies, and other centres of power began to take seriously the theories and affordances of networks, they started building the assumptions of those affordances into their systems as a means to re-exert control. As systems start operating on these principles, the world contorts around them to oblige.

40

The Network Turn - Part II Cultural Networks - 3 Culture Is Data

Nothing is naturally a network; rather, networks are an abstraction into which we squeeze the world. Nevertheless, almost anything can be turned into a network, whether it be the interactions between characters in Shakespeare’s plays, the dissemination of …

43

In such a context people go from being nodes to being the connecting edges between places, as in the article ‘A Network Framework of Cultural History’ (and the accompanying video), which provides an overview of the movements of humanity over the past 2,600…

46

Without the metaphorical and analytic power of networks, it is difficult to understand how the interactions of characters in Hamlet are in any way comparable to the dissemination of memes on Facebook, or the chemical compounds shared between food ingredients. What makes them compar- able is an intellectual and methodological shift by which we abstract our objects of study into data points that can be entered into a database or spreadsheet. This does not imply some shared property intrinsic to each of the subjects under study, rather it implies the widespread utility of networks as a lens through which to view many aspects of our shared world.

51

It is important to remember, however, that almost any network data set, even in the sciences, is incomplete in some sense.

52

While this of course will affect the results of any quantitative analysis, this incompleteness is less of a problem than it may seem at first, for three main reasons.

52

The first reason is that, particularly in the context of the humanities, the bias in the source data is often itself of interest to the scholar.

52

The second reason is that, even in an incomplete network, the relative importance of nodes and edges according to a given network measurement can still yield meaningful results.

53

The third and most important reason why the incompleteness of net- works does not pose an insurmountable problem is that the results of the quantitative analysis do not serve as a final result, but as a starting point for further, more detailed inquiry in the vein of traditional scholarship in the arts and humanities.

53

The Network Turn - Part II Cultural Networks - 4 Visual Networks

Even with this simple overview, it quickly becomes apparent how much the arrangement of nodes influences our reading of a network.

63

What is useful about Bertin’s work is that it gives us a point of reference for reading the rhetoric encoded in design decisions.

63

It is not only that computer graphics look precise, computer graphics software is encoded with assumptions about graphical meaning rooted in Bertin’s theory of graphics as monosemic and unambiguous.

64

But we do need to be aware of the assumptions encoded in the tools we use so that we can bend them to our own needs.

64

The force-directed layouts are a class of algorithms used frequently in Gephi and the visual form with which readers will be most familiar because of their increasing ubiquity through high-volume use of Gephi and similar tools. In a force-directed layout, each element of the network is modelled as though guided by unseen forces, much as in the physical world: gravity, electromagnetic repulsion, material elasticity. Nodes may be thought of as electrons on a two-dimensional plane, forcing each other apart as they approach one another, and the edges may be conceived of as springs physically anchored between two nodes, drawing them together even as they repulsively push away.

66

Tommaso Venturini, Mathieu Jacomy, and Pablo Jensen have begun to address this challenge by explicitly defining the strengths and weaknesses of what they have called visual network analysis (VNA) and providing guide- lines for how to make effective use of algorithmic layouts to expose the topological structure of large networks…

67

The argument underlying VNA is that understanding the conceptual and mathematical underpinnings of the layout algorithms and choosing appropriate settings can result in effective, legible graph spatialisation.

67

Though data analysis is widely considered synonymous with particular forms of visualisation these days, charts and graphs were not widely accepted until they were formalised in the mid- twentieth century when computing and, specifically, computational informa- tion processing was on the rise.

68

…visualisation is not better than narrative argument or mathema- tical equations for communicating ideas, but that it provides an additional means of producing, exploring, and analysing information that has proven value in both the liberal arts and the sciences…

70

Visualisations of networks are unsettling because the graphical language used to produce them is not, in fact, precise.

70

The Network Turn - Part III Manoeuvres - 5 Quantifying Culture

Numbers have the ability to capture certain attributes that cannot be gleaned simply by reading text or looking at images. Statistics can make an argument that cannot be expressed by words alone. Despite this, the quantitative is perceived as at odds with the normal practices and tropes of cultural commen- tary. Quantitative network analysis presents a challenge to traditional scholar- ship in the humanities not only in terms of methodology, but also in the ways in which we write about the findings from such a process.

74

…the use of any of these tools requires a prior mental manoeuvre of translating cultural artefacts into an abstracted form to see whether they are compatible with the input requirements of the available tools. This process of abstraction is not only a way of thinking, it also gives us algorithmic power.

75

Moreover, and as we argue in more detail in the following chapter, the research process will involve shifting back and forth across this continuum when addressing or honing a particular question.

79

…single off-the-shelf algorithms, such as the centrality measures mentioned earlier, can provide useful first insights into the data, but the likelihood of any single measure being a reasonable proxy for the specific cultural attributes or phenomena one is looking for is minimal. To find the nodes or individuals of particular interest, or to develop more nuanced research questions beyond structural network questions, the measures typically need to be combined with other quantitative or qualitative approaches.

83

…by looking at how the different categories of nodes interacted, they observed that the most prolific leaders frequently and repeatedly wrote to network sustainers, meaning that the shortest paths across the network were also the ones most frequently traversed by letters and, by implication, carriers.

84

Categories of people such as those defined for the Protestant network show how network analysis enables us to manoeuvre between different registers of quantitative analysis, from micro, to meso, to macro. Both the use of centrality measures and the construction of network ‘profiles’ are about generating a quantitative description of certain attributes of a single node as a function of their position in the entire network.

84

Centrality measures can also be used to think about the mesoscopic properties of complex networks.

84

In light of the changing ways in which culture is studied, we contend that a pressing duty is placed on the university to prepare future generations of academics by offering suitable combinations of courses in humanities subjects, programming, and statis- tical methods.

86

…how we can produce scholarship that is accessible to audiences from all the contributing disciplines. In constructing practices of research that lie at the intersection of several disciplines, we need to make sure that we have created something greater than the sum of its parts, rather than something lesser, such as an inter- disciplinary silo in which we are only speaking to other displaced scholars…

88

The Network Turn - Part III Manoeuvres - 6 Networking the ‘Divided Kingdom’

The correlative to the scales of investigation – the satellite view, the aerial view, and the archaeological dig – are the different kinds of argument that can be proffered.

95