Tài liệu Weaving Networks pdf

How to play whisper down the lane: buildingthe router the folks at BBN went head first into figuring out how to design a network of computers, and related things like how to pass messag

T A B L E O F C O N T E N T S

I NTRODUCTION 1

WHAT IS A NETWORK? 5

I NTERNET : A 900 LB M ODEM IN THE ROOM 6

F EATURES OF N ETWORKS 14

T HINKING A BOUT N ETWORKS 26

T HINKING A BOUT S YSTEMS 31

W HAT MAKES A N ET , W ORK ? 36

T HE L IFE AND D EATH OF N ETWORKS 47

NETWORKS IN ACTION 61

T HE K EYSTONE A DVANTAGE 63

T HE H IDDEN P OWER OF S OCIAL N ETWORKS 80

T HE T IPPING P OINT 86

N EVER E AT A LONE 93

T HE W ISDOM OF THE C ROWDS 96

E MERGENCE 100

W HO R ULES A MERICA ? 113

W RAP - UP /S UMMARY 116

STRATEGY: DESIGNING CLOTH 119

T O NETWORK OR TO NETWEAVE ? 120

J OIN ‘ EM : G ET I N W HERE Y OU F IT I N 121

B EAT ‘ EM : C LIMBING THE N ETWORK 127

J OIN US : I F Y OU B UILD I T T HEY W ILL C OME 136

TACTICS: CHOOSING THREADS 147

Y OU DOWN WITH IPP? 148

M APPING 158

M INING : D IGGING FOR G OLD 165

B UILDING AND S HAPING : K NITTING 168

M ANAGING N ETWORKS : D EPLOYMENT 182

S PREADING L IKE W ILDFIRE 194

W EAVING W EAVERS 201

ENGAGEMENT: WEAVING THE WEB 203

H UMAN B EING : A P HILOSOPHICAL P ORTRAIT 205

H UMAN D OING : A P RACTICAL P ORTRAIT 225

O N B EING AN ‘I’ 246

O NE -O N -O NE : D EALING WITH AN OTHER ‘I’ 255

NETWORKING TO NETWEAVING 292

A VOIDING NETWORKING T RAPS 294

S OCIETY ’ S GLUE 295

END 301

A FTERWORD 301

A PPENDIX A: SNA Q UESTIONNAIRE 308

T HE M ETA -M ODEL : Q UALITY I NFORMATION 310

I NDEX 311

B IBLIOGRAPHY 312

M ISCELLANEOUS 315

BOOK TITLES 318

Introduction

I think many other books on ‘networking’ fall short of giving you a complete picture of networks and networking I am writing this book These other books on networking do describe a few of the individual actions you take to be a good ‘networker.’ However, these books fail to explain how or why these activities work and how what you do fits into the larger context of the network In an analogy, other books teach you how to hook your computer to the internet, but do not tell you what the internet is!

Imagine someone sold you a book about driving, but never mentioned traffic lights, roads, and other drivers They only explain shifting, turning the wheel, accelerating and breaking That would be absurd, wouldn’t it? Would you feel comfortable getting on the road? Would you know how to plan routes to your destination from where you started?

To understand networks and networking, at least in the realm of computers, you have to get certifications and computer science degrees There is even a whole branch of science dedicated to researching the features of networks But most of these networking books say nothing about the properties of computer networks or even business and social networks They tend to slink by without discussing those things Send you off and wish you fun driving, lol (laugh out loud) Who could possibly pass up the opportunity to relate ‘networking’ to the rich metaphor of the internet?

The book you hold in your hands is not an introduction to networking as you know it, nor as you think about it This is not a feel good, tip-filled book that promises to having you ‘up and running’ or

‘making millions’ in a matter of weeks or months Nope This is that book about understanding what networks are and how to operate in them It is about understanding what you’re doing, and then doing it well

This book is about months and years of fun work, after all you will be making hundreds of friends along the way This is the kind of book where make mental notes of things, and write stuff in the margin that nobody else sees This is the book that will give you those eureka and

‘oh’ moments that explain what other books haven’t This is the kind of book that builds character first, intelligence at a good pace and your

money slowly Okay, there’s a little exaggeration in there, but you get the gist

What’s in a name

The first inkling I remember having about something deliberately weaving and uniting different things was in the novel Chick Pea by Isha Schwaller De Lubicz In this book, a young Egyptian boy was tutored

to watch how animals shuttled back and forth in the natural world He was to watch the animals for a few days to understand the interconnection between their activities and lifecycle

I’ve also picked up the name netweaver by browsing through a few blogs on the internet about Social Network Analysis and Organizational Network Analysis There are a few organizations popping up with the explicit focus on building networks by introducing people, businesses and organizations with like or complementary skills This kind of awareness and activity characterizes netweaving and the netweavers that

Beginnings

This section explains the reason for the book, the overall gist of the book, and how it came about

What’s all this Network Stuff?

In order to get you started networking like a master you have to know what a network is The hazy ideas that people walk around in their heads is not good enough This chapter conveys the structure of networks from the science of network, yep, there’s a whole science that studies networks In this chapter I put the science of networks into context as well as introduce you to the terminology of networks This will help you understand what’s in the rest of the book It will also give you a deeper understanding of networks than most others

Introduction

in business, the economy, teams, super-networkers and social classes The aim here is to give you multiple viewpoints and filters so that when you look at networks, you can notice a lot more of their subtle nuances than their basic structure

Designing a Cloth: Strategy

Having looked at basic network structure, and some real-world examples of networks, now you can start to plan what kind of networking you will be doing All networking is not the same; there are different purposes for which we network We will address the differences between these broad strategies, or stages, here

Choosing a Pattern: Tactics

Whichever strategy you choose, you will need a plan of action to execute that strategy Hence, we have a section on the tactics of networking This chapter sharpens your ability to detect what’s going

on in the networks around you This chapter also shows you how to approach and interact with actual networks

Weaving the Net: Engagement

Once you plan your work, you have to work your plan To excel at networking, you have to be a better than-average communicator This section shows you how to drastically improve your ability to listen and speak with impact This is where your networking plan meets the people in your networks, where the pedal meets the metal Being a good networker is not just about knowing people, but about building solid relationships with people This section teaches you to form these lasting relationships

We give the name ‘net’ to a group of strings knotted and/or woven together, in some pattern, for some purpose The purposes of nets range from catching fish, to keeping bugs out of our homes (screens, duh) We even use a kind of net in our kitchen to drain water out of spaghetti (colander), or sift flour into a fine powder (sifter) You’re even wearing nets (hopefully, especially if you’re in public) in the form of your clothes

We can think of cloth as a tightly-woven net of string not meant to catch (fish, bugs) or separate (flour, spaghetti) but to cover our bodies Curtains and drapes do not cover our bodies, but are woven to keep the sun out of our houses Nets have different purposes We make some nets are designed for their substance (cloth) me nets are designed for the spaces between their substances (strainers) People design other nets for the substances between their spaces (fishing nets)

We can understand a net as an interlacing of materials, for some purpose I could wax poetic about the nature of materials, density of the net, the space(s) between the materials and how different nets serve different purposes, but that’ll be beating a dead horse I talk about these physical nets because at some other points in the book this discussion will be directly applicable.1

1 Refer back to physical nets somewhere

But we’ll not be talking about physical nets We’ll talk about abstract networks The nets we are talking about are the ties and connections between people, businesses and organizations We’ll use the example of networks in the world of computers, biology, and other arenas to shed light on how people connect to each other

It would be completely boneheaded to think about networks without referring to the most famous one of all: the Internet Not only is looking at the internet obligatory, it is probably one of the only networks that we can actually trace its history and growth with any bit

of rigor We can’t do that with networks of epidemics, proteins or ideas with as much precision here we go I’ll look at the history and development of the internet

ROOM

A while back, some really smart people invented what we today know

as computers They were big, expensive (half a million apiece), took up whole rooms and basements in buildings For the most part, before the 1970’s only very large companies, the federal government and a few universities had computers

As manufacturers made computers faster, more flexible and cheaper,

a few more businesses, universities and government departments could and wanted to use them When computers were originally designed, they were relatively self-sufficient universes unto themselves Even two

of the same computers from the same company couldn’t talk to one another Nowadays to get your computer to work you almost have to

be connected to the internet My, how things have changed

Logins and Time-sharing

In the 60’s and 70’s, more and more universities bought computers for their researchers to use Many of these computers were bought with grants from the federal government to do research

With multiple researchers wanting and needing computer time, a problem arose that the computers, for the most part, could only do one set of things at a time If it was a gender, we could say computers were men As a researcher, in order to do your work, you had to be at the computer, putting in your data and giving the computer instructions, from dusk 'til dawn, and from start to finish This meant you had to work at odd hours to get your job or research done This was a pain

As more and more researchers gained access to computing it was necessary to have two things: the ability for multiple people to tell the computer to do things, and the ability for the computer to know who is telling it what to do me enlightened programmers created two things: the login and time-sharing

We are familiar with logins We do it with our computers and email all the time Before logins, a computer couldn’t separate the commands and files from one user to the next After logins, computers were able

to sort out and process different files from different people This made

it easy so when one researcher told the computer to delete something it didn’t delete another person’s files

Time-sharing is not as famous as logins, but just as important Back in those days, computers were remarkably slow We have faster computers even in low-end calculators when a researcher wanted to do something, it usually took a long time As computers could only do one thing at a time, one researcher had to do all of his or her calculations and programs Only when they were finished could another researcher run their programs and calculations

Logins and time-sharing enabled multiple people to work with a computer at the same time The computer was able to sort out the files

of each user and it could also keep the instructions of each user separate This enabled four or five people to log in, start and run a program, and come back after other people worked on the computer The problem solved is just like getting an education When you go to school, you do not have two months of math, then two months of English and so forth Early computers were designed this way, where users needed large blocks of time to get what they wanted done With time-sharing and, people were able to do what we experience as scheduling In a school schedule, you take part of one class, switch to another, and so on The process would still be slow, but researchers could do bits and pieces of their shtick over weeks instead of all-nighters days in a row

Proliferation of computers

As time went on, computers became cheaper as more were built, and more companies got into the game As the fields of computing and manufacturing were developing fast, each new computer model was almost one-of-a-kind, much like cars today Every year a new model comes out, the parts of the previous model do not work on the new car

What seems counter-intuitive today was that even computers that were the same model and from the same company couldn’t talk with one another! As time passed, some programmers were requesting that the computers they bought would be able to talk to one another Of course it would be frustrating to have two half-a-million dollar computers sitting next to each other but not able to talk Computers weren’t designed or programmed to talk to other computers, regardless

of distance The programmers sometimes literally had to learn a new programming language for each computer!

mewhere in there, after the development of time-sharing, manufacturers came up with what’s now called the ‘client-server’, back then the ‘master-slave’, relationship between computers Before, if you wanted to connect to a computer, you had to be sitting right in front of

it to work on it The master-slave relationship enabled a user to log into

a computer from various places through a remote computer terminal The big machine itself was the ‘master’ computer where all the computation and processing took place while the remote ‘slave’ computer terminal was more like an access point to the ‘master’ computer than a computer in its own right It is kind of how Google is not on your computer, but you can access its pages through the internet

I need a network

They say the decision to build the internet was spawned by Bob Taylor, head of the Advanced Research Projects Agency (ARPA) Taylor was upset that he had three different computer terminals in his Pentagon office He thought it was bone-headed to have to have three separate terminals to log into three different computers He wanted to

be able to only have one computer, and be able to log into all three of those computers

Parallel to his pet peeve, a lot of universities were also requesting money from the ARPA for more computers It seemed as if every university wanted their own computer Now, with the advances in time-sharing, he realized it would be a lot more cost-efficient if he could enable other researchers to use the computers that already existed What he envisioned was having one terminal to be able to log in to multiple remote computers This would be cheaper in the long run for the ARPA both because computers were real expensive, and because most universities didn’t use the full horsepower of their computers anyway

But to do this, he needed some way to connect all these different universities and researchers Roberts pitched to his boss, on the fly with

no preparation, the idea to build a network that enabled other researchers to plug into other computers, even if they didn’t have their own computers He got a cool million That was back in 1966, which translates to about $6 million in 2008 dollars Imagine that Building the core of the internet only cost $7 million!

Roberts told his boss that he knew it was feasible, stretching the truth

a little because he had already seen a demonstration of two computers talking with one another A year earlier, Tom Marill showed Roberts that it was possible by connecting two different computers using a full-duplex connection2 Marill set it up so that the computer would group messages, send the message and check to see whether the message arrived If it didn’t get a signal that the message arrived, it would send the message again He called this back-and-forth a ‘protocol’ as it sounded like the process of diplomacy

With this experience under his belt, Roberts thought that simply enabling a few other computers to do the same things with one another would be relatively easy

Creating a network

One of the early problems that stalled the creation of the original internet was that it would be real costly in man-hours The original proposal was to program all the computers to communicate with the other computers This demanded that each university somehow learn how to translate programs and instructions from one computer to another, much like Marill did This was costly because each university would have to pay its programmers to do this Aside from that, when computers are programmed to share time, the computer actually has to spend some of its own power to think about who should get what time Adding another function would also add another layer of things for the computer to do and reducing the percentage of the computer dedicated

to actual work Even more, most of the universities who already owned computers didn’t really need to talk to other computers Remember, the idea was for other new researchers to log into the existing computers All in all, the universities they were against the idea of such a network

2 Full duplex means that both computers can send and receive messages at the same time (like phones) while half-duplex means only one can send at a time (walkie-talkies)

All this resistance was discovered during the first meeting at Ann Arbor where Taylor and Larry Roberts, who Taylor brought in to be in charge of building the network, of ARPA pitched the idea of such a network to the universities As you can see things were borderline disastrous During the meeting one attendee, Wes Clark, passed up a note that said “you have the network inside out.” Intrigued, Roberts and Taylor drove Clark to the airport so learn more about what he had

to say Clark suggested that they instead of building a network where each of the computers had to do all the translating themselves, to build

a network of smaller more specialized computers would handle all the translation and traffic between the computers We now call the network between these specialized computers ‘routers’

and the network that they exist on a ‘subnet.’

This idea solved a few critical problems No longer did the host universities have to figure out how to translate from one computer language to the next, they would only have to figure out how to talk to the specialized computer The specialized computers themselves could

be programmed to handle and route the traffic among themselves more easily as they would all speak the same language This also enabled the network be directed under the control of one group, ARPA or whoever won the bid to build the network, instead of the smattering of universities and researchers Can you say eureka!

After going back to the drawing board and following through with this idea, the ARPA folks pitched the idea to the same audience who was more receptive to having to do less work After the second meeting, ARPA put out a request for a proposal (RFP) to companies across the nation to design and build a network that enabled researchers from any of the major universities to log into and use one another’s computers The bid was to create a network of “interface message processors” or IMP’s for short This was especially dicey as most of the requirements and questions surrounding how to build such a network theoretical or new

Despite what we think about crony politics and ‘pay-for-play’, the bid was won by the company that had the smartest people and the most thoroughly researched idea on how to put the network together Its research in preparing its proposal for the RFP alone cost it $100,000 That company was BBN, standing for Bolt, Baranek and Newman In its heyday BBN was called Boston’s “third university” since it hired mostly dropouts from Harvard and MIT No wonder they were smart enough to figure it out

How to play whisper down the lane: buildingthe router

the folks at BBN went head first into figuring out how to design a network of computers, and related things like how to pass messages between them, how to talk to the other computers, you know, minor details A lot of the big questions they answered in the course of preparing their proposal, so they could get right to work on the nitty-gritty

One of the critical aspects of this network, if it was to work, was how

to relay messages from one station to the next There was actually a large theoretical battle between Paul Baran3 and AT&T about the feasibility of what we now know as digital networks AT&T was against

even the possibility of digital networks as its own equipment at the time

was all-analog You think moving from analog to digital television was something? Imagine how hard it was, or would be, for the whole phone

3 Reproduce the network diagrams

Baran figured that the regular hub-and-spoke centralized networks were very susceptible to breaking down if attacked All anyone had to

do was take out the core, and everything else was left isolated He thought that a decentralized network wasn’t much better as those networks though had more hubs, could be taken out with just a few key nodes being attacked He came up with the idea and advocated for a distributed network where each node on the network connected to three or four other nodes This way, if say ten or twelve nodes were taken out, the rest of the network could still pass messages to each other

The folks at BBN took his work seriously and went to work figuring out how to translate Baran’s theoretical arguments into actually programming their computers They accomplished this by breaking long messages into smaller messages, similar to mailing bricks across country to build a house Those messages were called ‘packets’ on account of an Englishmen named Donald Davies who thought about the same problem It would be the job of the computers to know which routes to send the messages (bricks) through to get to the ultimate destination Each computer would be equipped with a table, called a routing table that enabled it to figure out the best path to travel

to the destination

Machine(s)

the original IMP was a Honeywell 516 The IMP-guys went to work redesigning parts of it, and adding on other things that will allow the IMP to talk with other IMPs and other computers It was the job of the different universities to come up with a way to connect their computers

to the specifications of the IMP We would call this specification of what to do, how much and when an API, which is short for application programming interface

Now this connection had to work on two levels On the hardware level the computers had to decide on a common cable connection, this

is where Ethernet cable was created A more subtle feature here would

be the level of voltage running on the wire, little details like that In the world of software, these computers had to be able to talk with one another in the language of 1’s and 0’s and not any one computer’s hardware language This allowed each computer to talk nice, but it was then hard to figure out where one part of the message started (say the address or header) and the other part of the message (actual message or body) stopped Standardizing these things and distributing them to the sites slated to plug into the network was critical

The invention of the internet

At first, when this network was created it only connected computers from domestic universities, government agencies and BBN At the same time, a bit of work was being done across the big pond (the Atlantic Ocean) by the British and others Eventually the idea came about to connect the US network to other networks, hence the term internet, short for either inter-network or international-network

When they hooked these computers to one another, they called them

a network Later, as more and more computers wanted to get into the little networks from different departments, another problem arose When twenty users from different computers were trying to pass messages to specific computers over the network (through a broadcasting hub), and the bandwidth (rate of information exchange) was very low, how do you make it so that the network does not get clogged with non-essential information? Put another way: how do you make it so that computers from one department can talk with computers from another department, without disrupting still another department? Routers!

routers act as a drawbridge or stoplight between smaller networks (or subnets), they let some traffic through, while stopping other traffic They handle and route information intended to get to other networks Routers work like your post office If you want to mail something to someone within three or four blocks, you go put it in their door When the intended recipient resides far away, you use your post office, complete with address, return information and postage

When you can put mail in your neighbor’s door, we call that area a neighborhood When you can send messages to others computers directly, we call that area a subnet, or an intranet As intranets multiply,

we call the connectivity between multiple intranets an internet And

obviously, when we internetwork multiple internets, we call it ‘the Internet’, as opposed to ‘an internet’

[Nodes, links, hubs, routers, shapes, internet]

We are done with the time-lapse history lesson on the birth of the internet Now we are going to comb through the information to explain some of the details of networks and networking I am going to throw a little more jargon your way so you can understand what’s what me of this information will be known to you already in bits and pieces, this section is here to weave those into a tapestry (a pictorial wall-hung rug)

Nodes and links: the patterns that

connect

In a network, what is linked? Nodes Node is the word we use for points in the network In society, a node is a person In business, we can think of business departments in a business as nodes We can also think of businesses themselves as nodes in a network of businesses (suppliers and customers) be careful here, when people look at a

‘business network’ they can mean either people and departments within

a business or they can refer to multiple businesses themselves To keep things clear, when I talk of a business network I’ll be referring to people within a business as nodes When I say a network of businesses I’ll be referring to whole businesses as nodes in the network But because the idea of a network of businesses is relatively new, there is no special terminology to make it distinct from a business network

Strong and weak ties

In talking about networks, we focus for the most part not on individual nodes, but how they relate to other nodes In this, network researchers measure how stable a relationship is between two nodes The links between nodes that are stable or close relationships are called

‘strong’ ties, while the links between nodes that are not very stable or closer are called ‘weak’ ties

You and your boss may have strong ties at work, since your jobs complement one another, but loose ties socially, because they are an ass A business may have a tightly knit relationship with their suppliers, high communication, a lot of collaboration, but weak ties with their customers, bad customer service, products that do not get updated etc

me relationships are reciprocal, or symbiotic if you want to use a biology word, like Microsoft and Intel Microsoft needs Intel microchips to run computers, and Intel needs Microsoft to sell good software for people to use That would be an example of a strong, or tight, link But with the rise of AMD which rivals Intel Microsoft’s relationship with Intel does not necessitate such a strong tie, but cutting that tie would be foolish nonetheless

In contrast, an example of a loose tie would be me loosely tied to a corner store, which sells the wonderful Tastykake pastries that I love I can buy Tastykakes at a lot of places, and I am not tied to any one place

to get them Just the same, the corner store has many customers, and

do not rely on me for a lot of their business

There are other examples such as in biology where we may have two species that rely on each other, such as plants needing bees to pollinate them, while bees needing the nectar from the plants for fuel, as examples of strong ties By contrast, two prides of lions living far from each other may have loose genetic ties to each other

what we have here so far are nodes, individual points in a network, and the strong or weak links that connect two or more nodes Obviously, in any network, each node has many links, and if you look

on many levels, each node may have thousands of links to different parts of its environment For example the cellular networks within an organism are composed of millions of strong ties when we talk of nodes in a network, we are selecting from a variety of levels of possible networks and focusing on one when we talk of a node, we are restricting ourselves to talk of that node in relation to similar kinds of nodes For now, I say only that a network is only one level of a system, and we can view any node as a member of different kinds of networks

I will give examples and go into more detail in a later section

far we have been talking about relatively simple networks with two players Now we are going to throw in a couple more nodes into the picture For instance, let’s say there are you, me and Jerry Maybe you and Jerry are close, I and you are close but me and Jerry do not get along I am pretty sure you can apply weak and strong ties appropriately

in that example But what happens when there are fifty nodes in a network?

Dense and loose networks: crowds and groupsWhen you move to talking about a lot of nodes in a network, researchers think less and less about strong and weak ties and start

thinking about the properties of the whole network If in a group of 50 people only ten or so know each other, we would call that a ’loose’ network If say about 35 people knew other people that would be a more ‘dense’ network That is dense in terms of metal vs cotton candy, not dense in terms of not being bright

The shift from node-to-node analysis to the network view is like the shift from ‘I and’ you to ‘we and them.’ The shift to analyzing a network moves from looking at the quality of individual relationships to the number of relationships or links between the nodes in the network

We actually talk about this rather naturally in English We call loose networks, that do not have a lot of connections, a crowd We talk about

a dense network, where ‘everybody knows your name’, as a group

We use the word crowd to refer to collections of people that have no lasting ties to one another When we say group, we talk about collections of people who have relationships with one another, loose, tight, fun, miserable, business, volunteer, or any other kind of relationship Crowds turn into groups over time as they interact with each other, get to know one another and make more and more connections There’s no hard and fast rule as to when a crowd becomes

a group, but in the world of network theory they are trying to find some kind of magic number to figure that out My little rule of thumb is that when more than half the people can identify everyone it is a group, anything under that we have a crowd

In observing networks, this becomes important as networks exist as collections of nodes that have links that exist over time These links enable each node to identify itself, the network(s) it exists in and other nodes

Anybody will confirm and talk about the difference between a crowd

of people at a Philadelphia Phillies World Championship (woohoo) game versus a high school or family reunion We could call them both crowds, but the distinction group emerges when you have multiple connections between people, especially strong connections To be a little precise, when I talk about crowds I will use the word people and when I talk about groups I’ll use the word persons This is similar between the distinction cows and cattle I use the word persons as I focus your attention on the fact that nodes in the network can identify each other as distinct nodes and not just ‘some other node’

when I was talking about my rule of thumb for the distinction between a crowd and a group, the rule of thumb measures the density

of the network When we look at networks, we look at density first and strength second Network-based thinking looks first at the holistic view

or forest and then the connections between individual nodes or trees how does a crowd become a group? If we take a newly forming organization, be it a college fraternity, a school club or a business, at the outset we have a crowd of people in a room, or at a meeting, who probably do not know that much about one another In team building exercises just about the first thing done is everyone introduces themselves and tells something about themselves This is usually the first stone on the long road to group Using my rule of thumb, when more than half the crowd can recognize and perhaps name the other individuals it becomes a group

The fact that the members of the crowd/group can identify with each other is only half the battle The other half of the battle is noticing and measuring how connected the members of the network are to one another As they interact and sort out who will do what job, or fill what position in the organization, we note that each member of the group starts to recognize the other members as individuals filling the various roles of the group

In some groups, like organizations and businesses, the kind of connection that gets made can be even more important than the fact of the connection itself In some groups, like businesses and organizations, humans create reporting structures, responsibilities tied to roles, not the individuals These reporting structures try to formalize some of the aspects of the kinds of connections (authority, responsibility) of different roles in the group However, these formal distinctions do not always capture more subtle aspects such as mentor-mentee relationship, social adversaries and topic experts the organizational chart is not the end-all-be all of how a group works, we’ll see more about this later

In less organized contexts, you might have roles like the ‘life of the party,’ the ‘nosy person’ ‘the gossip’ and others The kinds of ties in a social network can range from confidant to party organizer to mortal enemy

While we start to become aware that there are different strengths of connectivity between nodes and different densities of connections within a network, we also start to see the limits or boundaries of them

Within these boundaries, some of the nodes may have more or less connections to other nodes than others In nearly every context, you can find someone who knows just about everyone

As networks evolve, they gather more connections and interconnections among them We call this attribute density If you take out of a ten-person network, the person that knows everybody, the one with the most connections, the network becomes much less dense, or less interconnected A dense network usually has redundant and overlapping ties An example can be a network of roads getting me from one place to another If I wanted to get from node A to node Z I could use a couple of different routes, not just one

Networks that have a lot of overlap and redundancy we call dense, tightly-knit or tightly-woven networks while we talk about loose, loose-coupled or loose-fitting networks when there are not a lot of connections and overlap In the travel example we could think of intersections as nodes If a road network has a lot of intersections, it is a dense network When we shift from viewing individual streets to connections between cities, the number of connections drastically reduces Also in this example we could say that the strength of a connection between two cities or intersections is measured by the traffic capacity of the street or road A four lane highway indicates a stronger connection between two points than a one-lane dirt road

Think about your networks of friends You might have a few isolated groups of friends that mostly know each other Within each of these groups, we would describe them as a dense or tight-coupled network But, when you have two groups or ‘circles’ of friends that are not very connected to each other, perhaps with only one or two people being in both groups (like you, duh) then we would call that loose coupling the phrases tight and loose coupling can also describe the density between networks also

If you meet with someone once or twice, you probably wouldn’t say they are in your network (LinkedIn notwithstanding) But, if you see them at work and on the weekends pretty often, we could start to say they are in your work and social networks We would probably also say that they are a strong tie when compared to other people at work whom you do not go out with on the weekends or after work What I want to stress here is that the existence or strength of a tie is in relation

to the network within which you measure it

All kinds of disciplines apply the notion of loose and tight coupling to whatever they study Biologists talk about the loose and tight coupling

of species in an ecosystem In business, supply chain analysts constantly think about how important suppliers and customers are to the creation

of their product Social workers and child advocates try to reduce negative loose and tight-coupling, and encourage positive tight-coupling between their clients and productive role models, groups and organizations in their client’s community

we just took a look at the features of individual links or ties (strong and weak) and features of a network as a whole (dense/tight and loose) Let us look at another aspect of networks

In the development of a group, some people will be more important than others When you look at the roles and responsibilities in an organization, this may be the president, chair, etc That is not our concern in network analysis When we think ‘important’, we think connected Instead of measuring ‘importance’ by what responsibilities and power, we measure centrality, how many other nodes a node is connected The most connected or central nodes we call hubs

Hubs: Where Everybody Knows your Name

Hubs are nodes, like squared Hubs are nodes with lots of connections Well, more connections in a network compared to other nodes if, on average, any group or random crowd of 50 people have an average of let us say 5 links, well, hubs would be the ultra-popular people who have like 15 links to other nodes Hubs are the popular kids

The most connected nodes we call hubs Most airlines have hubs, specific airports that handle a lot of traffic A lot of time, when you fly and have a layover, you are likely to be at one of the hubs for that airline Airlines send most of their traffic through their hubs, and therefore the hubs have the most connecting flights to other airports The same thing happens with busses, trains and communication networks

In computing, it is a little different Hubs are pieces of equipment that listen to and broadcast almost all of messages from and to the other computers it connects It eavesdrops on all the computers and puts what it hears on a loudspeaker to the other computers Website hubs are still a little different Website hubs have many incoming and outgoing links to one another, but they are not designed to broadcast

every message, they are more like airline hubs, which are designed to route traffic to specific destinations

Hubs arise naturally in the nature of any kind of network There are a few popular books about networks that talk about hubs like LINKED by Albert-Lazlo Barabasi, NEXUS by Duncan Watts and other books on networks that you might find interesting to read after you finish this book These books look at networks from the world of theory and research These books look at the general properties of networks, from biological ecosystems to the internet If you want to look into the mathematical and technical properties of networks then go grab them from your local library or used bookstore

Anyhoo, at the center of any group, organization or network, there is

a hub, maybe even a few of them Actually, in just about any groups there are a few hubs, just like there’s never only one popular kid, there are cliques of popular kids A hub in a business may be the president of the organization by virtue of their position, having to talk to many people from different departments or projects Less formally the hub might be the guy that everyone likes to talk to and who knows everyone and gets stuff done, so obviously if you have a problem, you go to them A hub could also be a small group of people, maybe there is a workgroup in a business, or a lunch table at school where many of the connected people talk with each other when people use the word hub they can talk about a physical location, a person or a group of people The common wisdom in business is that secretaries and assistants are hubs, because anyone who wants to talk to the boss has to go through them And because they know who the boss calls and who’s calling the boss, they know who’s who and what’s what They have their pulse on what goes on in the business

In physical nets, there are not really ‘hubs’ per se If you look at a fishing-net or a screen, there’s no cluster or clump of string or wire If there is, you have to untangle it for the net to work! People tend to make physical nets evenly spaced out, because they work better for that particular function But in our kind of networks, tangling is actually good for the network

Clusters: groups of hubs

In the social, computing, biological and business networks we talk about, stuff clusters, and that is good for the network Imagine if your stomach was spread all over your body Food would have to travel all over the surface of your skin to get digested The spread out nature may

be good for the protective layer of the skin, but not very efficient for digestion Good thing that evolution Right? For the most part, in our kind of networks, tangling and clustering is good, it represents density This density enables information, energy and power to move quickly throughout the network

The stuff that clusters and tangles in hubs can also be varied In addition to people clustering, along with those people information, power, knowledge about the network and many more things cluster

me clusters focus on who you know (social networks) while other networks focus on what you can do (business) or what you know (academia) whenever you have a cluster, or a network of hubs, much

of this stuff starts to accumulate in the mix of the cluster As the nodes

in the midst of the cluster are more connected to the other more connected nodes, they have more power and influence than other nodes

Now, it just so happens that the cluster in or hub of a network also influences the character of the network We define business industries

by the clusters of businesses (healthcare, marketing, carpentry), and social networks (by school, region, economic class, etc) by the kind of people in the midst of it all Hubs or clusters contain a dense network

of people who are connected and talk with each other all the time Due

to this, it is almost inevitable that their frequent interaction affects those people or nodes more than the less frequent interaction with the rest of the group The frequent communication among members of a network tends to create general attitudes and dispositions

Boundary spanners

People do not exist only in networks and clusters either They exist in multiple networks You have family, friends, neighbors, co-workers, exes and so on, so you live in multiple networks me people, you could say, specialize in being members of many different kinds of networks These people belong not to one particular group, but many groups Think of the kids that played sports, an instrument and were in science club However rare they were, they had lots of connections, but the important thing was that they had lots of connections to lots of different groups

William Domhoff, who we’ll see later, researches the social networks

of the rich folks in America He tracks the connections between people

on corporate boards, the schools they attended, the clubs they are in and the foundations they participate in He found that though there is a

lot of connection, a special few people still had many more connections than others

What he found is that among corporate boards, most people were on one or two boards But there were some people who were on three or more boards He noted that the people who were on three or more boards exerted a special kind of influence that others didn’t What he said was that the decisions that a person on multiple boards would make were based on what is good for more than just that one board When you have multiple people making decisions for the good of multiple companies, not just individual companies, then you have a tendency for those people to talk to others and make decisions that are supposed to help all the members of the groups It is the boundary-spanners that help form the glue of social classes, and large groups Other types of boundary spanners are people who participate in multiple clubs during high school, or parents whose kids go to different schools and participate in multiple schools The presence of lots of non-overlapping groups creates a more generalized identification with a group

Now, we see the downside of this with the fall of Wall Street financial firms Their insulated chumminess had them think they were invincible and insulated from risk Domhoff didn’t say that this tendency was good or always turned out for the benefit of everyone, he just pointed out how it worked He said that boundary spanners influence the group

by bringing ‘perspective’ that includes multiple groups, not just the current group This is also how some groups tend to avoid conflict

in our little hierarchy we start with individual nodes, adding more nodes and throwing in a couple links we get a network In the middle of any sizeable network of linked nodes, we usually find a few nodes that have many more connections than other nodes, and we call those nodes hubs If the network is big enough, the network might have an inner or sub network of hubs, which we’ll call a cluster The cluster is made up of hubs Though someone on a corporate board is technically

a node, they get on boards by being highly connected and respected in the first place The members of multiple corporate boards, which are clusters in their own right, are a hub’s hub

The hubs of computer networks are designed to simply parrot and broadcast any and every message it hears The hubs and clusters of

social networks do not do that, they interpret and relay messages to the

appropriate person or people, or if they want to stop the communication they do not relay the message In computer networks the function of broadcasting messages and routing messages (or not) used to be two different pieces of equipment, now it is just one- the router In social networks, both of these functions, telling everyone and telling a select few, are also located in one person, actually every person But let’s look at this second feature of hubs, routing

Routers: who gets what message

The nature of the presidential role in an organization demands that they get information and reports from different and disparate parts of the organization In effect, they and the committee or board becomes the hub or ‘brain’ of the organization They get their information from other positions, roles and/or committees In response, the president or board decides a path for the organization, and tells the rest of the organization to align with the new activity and direction Information, knowledge, power and energy clusters around the presidential role.4

mething subtle and special goes on here Unlike a computer hub, with its broadcast messages to everyone, in this business network, information, power and energy are sent in specific directions We have

‘communication channels’ and ‘chains of command’ and such Embedded within most organizations, there are specific routes that information must take (reporting), places where decisions are made (committee meetings) and ways to distribute the decisions (general body meetings, seminars, emails, announcements)

The rules of many organizations dictate that the president and boards and committees meet every so often, and report to the president In this way, we can say that information is deliberately routed toward the presidents We could also say that decisions are made by and routed from the president to the committees and then the general body And

of course, decision-making power in the role of a CEO or president of

an organization is the source of the power in that organization people

in these central positions have access to more and different kinds of information and therefore wield more power than other people in the network

A computer hub can be understood as nothing but a glorified repeater Computer hubs simply parrot and broadcast the messages that other nodes in the network have broadcast to it By contrast, a router

4 Router vs hub: between networks, within networks

routes messages to specific networks and destinations according to the address on the message The post office does not give your letter to anyone does it? When a computer router receives a message (since most of the time that message has an addresses or destination packet in it) the router figures out where it is in the network and how to get your message to its intended destination Just the same, a person in an organization can ask “I am on this committee, and this other committee needs this information, how do I get it there?”

In an organization, the routing of information and decisions are built into the rules and reporting structure of the organization Post office procedures tell each office and hub where to send messages to make the shortest number of hops (which trains, airports, trucks, etc) Computer routers have their own routing tables with the same kind of information, info about where other networks are Airlines and freight companies have whole departments and complex software to determine which routes its planes and railcars travel to get from one destination to the next

Air traffic controllers are the part of an airport that channels planes to and from the airport and tells the airplanes which landing strips to use

to get in and out of the airport In a train station, there’s a similar thing going on, but we do not usually know the name of the group or person that decides which tracks and so forth But in the case of the train station, there are literal switches that direct the train from one track to another, albeit guided by someone or some routing software All of these networks and systems require some software program, group or person to decide exactly what route each plane, train or message takes

to get across the network and between sub-networks

routers ask themselves “does this message go on this network” and if it does not “how do I get it to the other network.” If that router connects directly to the other network, it simply sends the message using the agreed to protocol If it is not directly connected to the other network,

it looks at its routing table to find out which router is, and send the message to them

the modem-router that you have at your house serves that function Not many people set up true in-home networks, but you can If you wanted to share photos or files directly between the computers on your wireless or wired computers you could You would have to make your computer aware of the other computers on your network Determining what computers are on the network is the job of your router

If you want to share information with your neighbor, for the most part you can get on their wireless signal If they live too far, you have to route through the internet for it to work (unless you want to bury some Ethernet cable between your houses) These routers are just like what I talked about before with the bridge-spanners They are the nodes in a network that participate in multiple networks, and negotiate the communication between them

You can think of the president of an organization or business as a bridge-spanner Though she or he is within the network, they communicate with the different committees and departments throughout the corporation They sit atop their business or organizational network

In social networks, the roles of switch and hub are not as defined or structured, so there is a less formal consensus-building process that goes on in deciding what to do and think, or not Actually, depending

on the type of information, and what part of the network you look at, each person in a social network can be a hub, a router or both This is why sometimes it is hard to identify in a social network who to go to in

a network to get something done because a lot of people know the pieces of the puzzle but do not make the decisions!

To negotiate where you fit it, you have to be able to follow the rules

of the network For instance, to drive legally, you have to pass a written test about the rules of the road, and perform a driving test to acknowledge that you know how to handle a car At the train station, there are lights just like on the road, but the controllers of the train

station are also in radio communication with the train conductors to exchange information about routing, timing and so forth The same thing goes with air traffic It even happens on the roads with road signs for general traffic and turn signals (sometimes) as communication between entities on the network

there are some other features of networks that are relevant and will

at first glance seem obvious However, what’s obvious does not exactly mean that the reason it exists is obvious we’ll get into the theory of networks to start to explain the technical aspects of the growth and structure of networks What we’ll be talking about is how networks come to exist, versus not forming at all

One of the main questions in network theory is how do they happen? What causes networks to be created versus having lots of independent nodes? Well, if they nodes are independent they are not really a node because a node is something inside of a network, you get the gist To answer this, lots of mathematicians interested in networks create computer simulations and run statistics on the data in the real world to find this out

What they are searching for is some kind of principle or number that says ‘a network will grow if X and Y happen.’ Let’s say we have 100 nodes And over time each of these nodes create links with other nodes Also over time, you can expect that the links they create also get severed due to distance, time, lack of use etc To create a network out

of these 100 nodes, the rate at which new links are made of course has

to be more than the rate at which old links are lost That is first Apparently this is a fundamental fact to get the network started in the first place, and the features of the network rely on another couple features

In this example, links are randomly created Any one or two nodes can be connected in some kind of random order But they’ve found that networks do not grow randomly, that is only in computer simulations On the internet, two websites are not arbitrarily connected

to one another Nope me writer or webmaster deliberately puts in a link to another website, mostly because that other website has some useful and relevant information Repeat this process with 20 websites and what you will get are clusters of links that are relevant One cluster

of links may be focused on health and the other may be focused on

appliance repair Whatever the reasoning behind the linking, clusters are not random They are based on what Duncan Watts calls ‘affiliation’ and Barabasi calls ‘preference’

Let’s balance this back with our random network idea now you still have some random links (only random because we do not know why they linked) and some targeted links Among these 100 nodes, you have two main clusters of about 10 links apiece Because these links are deliberate, you would expect that they last longer than the random links This is another feature that makes the organization of a network predictable: long-lasting links The network cluster can be created when the links in the cluster last longer than most other links, which helps get the cluster further above the ‘making new links faster than breaking old links’ threshold

We are about to do something a little weird We are going to throw relevance back out the window It was only a tool to get us where we are I only used it to create some kind of explanation as to why some clusters were made and others weren’t Relevance is definitely a necessary feature of real-world networks, but we are talking about random and semi-random networks for now We’ll get to the real world soon enough

the situation in our little 100-node world is that we have some links being randomly created and randomly dying Let’s say the rule is that any node that has more than one link will tend to retain those links longer than other links, i.e the links will dissipate much slower This makes it more likely that the nodes linked to a multiple linked node will also become a link with multiple nodes, as it holds onto one node longer, and can therefore have more time to pick up another link See!

We didn’t need relevance after all This new rule I made up is not that arbitrary, because as we know people with lots of friends tend to be good at both making new and keeping old friends around the rule is not so arbitrary after all

Now, here’s an interesting feature of that little cluster there’s a node that may be connected to two or more other nodes As each of these nodes keeps its links longer (let’s say 10% longer per link), it can build more and more links Eventually it reaches a point where it makes new links twice as fast as it drops links And of course, its partners will also have this capability, and therefore build their own links what turned out

to be three nodes originally, grows dramatically based on a small bias toward nodes that maintain links longer than others

There’s a weird fact of nature, observed in anthropology and business, that humans can only handle knowing about 150 people Any more than that and our world gets a little hazardous let’s throw this into the mix of our semi-random network the new rule is that any one node can only be connected to five nodes We will see that this makes very little difference to the building of clusters, or a super cluster

after the first node, A, connects with two nodes, B and C, let’s say that A and B each pick up another node now A has three links, B has two and C has one Now, A will be more likely to pick up yet another node than B or C But let’s say that in the next go round B links to C Now we have somewhat of a spiral Each of these nodes is connected

to one another, forming a triangle or triad Now, each of these now have two stable connections, well stable relative to the rest of the randomly created and destroyed links Because of this, these three become a cluster unto themselves Now eventually, each of these will max out at five links But because the links are not permanent, at least

in our example, one of the links will be severed

Even though the link between say A and B will be severed, they are at the nucleus of the group of links whatever new link they get, will immediately get linked right into the nucleus of the group Now let’s add something else The rule before was that any node that has more than one link will increase the time it retains those links by 10% let’s add another rule Any node that is linked to a node with more than two links will retain its links for 5% longer In this scenario, the node newly connected to either A or B (that just lost the link between them) will be inclined to get and retain more links also

Though we have been focusing mostly on A, B and C, all the links linked to them will be going through the same thing They’ll be bringing more links into the mix than before in this semi-hypothetical scenario

we can see that any preference toward retaining links longer than others will create networks and clusters out of a random bunch of nodes Now, there’s a lot of math behind figuring out at what threshold networks form, clusters form, networks and the like But we are not worried about that What we are concerned with is simply the birth and growth of networks from, well, no links

In the real world, researchers have found that links are also not randomly generated There is a bit of preferential treatment in there links tend to get made with nodes that are already linked In human society, it is more likely you become friends with someone who has a

lot of friends than not In our example we could say that nodes with a link are 5% more likely to pick up a new link This would capture what

is alternately called ‘affiliation’ or ‘preference’

In the world of network research, they have identified a type of network called a scale-free network Basically, in any network, there will always be more less-connected nodes than highly-connected nodes And the ‘scale-free’ comes into playing that says even in cases where you have highly connected nodes, there will be nodes within that group that are way more highly-connected than others, and within that group, etc

Thinking about scale-free networks reminds me of something I read

in the Philadelphia newspaper about when the Falcons drafted Michael Vick into the NFL The sportswriter was saying that we needed a new metaphor, calling him ‘superman’ wouldn’t work he said we even needed something better than “he’s the fastest kid on the block And,

he is still the fastest fasted kid on the block when we get all the neighbor’s blocks together to race He’s even faster than the fastest kid

on the block when we get the fastest kids who are faster than the fastest kids on the neighbor’s blocks He’s even faster ” you get it

In scale-free networks, there will always be a few highly connected nodes Now, in our example it seems as if we couldn’t get to this because I put a limit on a node only able to be connected to 5 other nodes If you think about it, that node knows a whopping 5% of that 100-node universe We know from Google, whose search database is connected to at least 60% of the internet that it is pretty much at the top of the food chain Search engines are at the top of the food chain, newspaper and news sites come next, then topic-based sites, blogs and commerce sites are somewhere at the bottom There’s a lot of overlap

in this The point is that our example didn’t capture the ability of some nodes (a person, business, group, organization, protein, intersection, etc) to be more connected to other highly-connected nodes Our example limited this But to recap, here are some of the general rules that move a random group of nodes to a network:

• New links must be made faster than broken

• Nodes with links tend to pick up more links

• Nodes with multiple links tend to retain links for longer

We can say that these are the three basic rules or conditions that must

be true for a network to form I know that you, as a budding networker,

can probably see that these are some of the rules behind what other books tell you to do: get out there and make new connections, learn how to get referrals, make sure you keep in contact with the people you meet Part of the problem is that some of the books on networking focus on one of the rules, not all of them while these books follow these three basic rules of networks, they are not giving you the whole picture

me things missing from these rules, but shapes the fields on which these rules are obeyed, are density and size A network can’t arise if there are too many nodes and not enough connections The rate at which links are made and maintained has to be related to how big the numbers of nodes are if a network is to be born

Shapes Network Members Make

Okay, that is a nice introduction to the basic features of a network

We defined nodes, strong and weak ties, density, hubs, routers, loose and tight coupling and a couple other things That couple other thing included the three basic rules of building networks that move a random group of nodes to a network Just ahead, we cover some of the more advanced features of networks and systems Admittedly, almost none

of the popular books I have read on networking have anything close to

a conversation about what is in the next chapter, so I will be making a whole lot of it up as I go along Actually, I’ll be pulling most of this

from books not about networks but about systems The intent here is

to help you understand the environment on which networks exist What I present will give you a depth of information that these other books take for granted, I am just making it explicit I am pretty smart so

I won’t mess it up, knock on wood

I intend to include information and insights from systems thinking, which I find to be closely related to thinking about networks The study

of systems is a great and fascinating discipline, but you are not reading this book to learn about systems I’ll keep it short and sweet I am including this because talk of systems parallels some of the talk of networks I want to give you a taste of how systems differ and contain similarities to networks I also wanted to admit that there are parts of this book that fall more into the systems-view side of things, but they are so subtle that you might not notice them, so I left them in

What is a system? And what’s the difference between a system and a network? We could define system as a network of different kinds of things interacting with each other You can think of a network as a layer

of a larger system We could say that a network is composed of similar parts in a system For instance, the Internet is a network of computers, but it is only a layer of the whole computing system What most people think of as the Internet relies on hundreds of different kind of chips, routers, protocols etc We sometimes mention a network of railroad tracks or roads, and those are a layer of the transportation system The network of track sits on concrete, gravel, wood and earth to give them a foundation

The definition here is not supposed to be precise, because there are many grey areas I want to stress that networks do not happen in-and-

of themselves, and it enhances our understanding of networks to understand the context or environment in which they exist We’ll also note that some of the features of systems influence and direct the coming together and falling apart of networks

5 Distinction: systems are networks that have api’s for other networks (social network

vs social system) – this distinction has to go after the keystone advantage section

me Basics of Complexities

I like an oxymoron every once in a while, so shoot me

Up until now, we have been looking at a taken-apart car and pointing out pieces Now, we have to talk about the car when it is put together

we have to put humpty-dumpty back together again and look at it as a living egg, not just a pre- omelet mess when we look at networks, they are not static, they are dynamic almost living things They grow and shrink over time, and change shape

There are three or four features of a system that define it as a living system Notice I said system, not network Also notice that I said living system, not just system This section will discuss what’s required for something to qualify as a living system, and you will see that much of it will be applicable to the workings of networks The two properties are that a system is self-refueling and self-propagating and habitat preservation Another closely related quality is development We’ll look

at each of these in turn

Self-refueling

Feeding yourself is the first sign of maturity It could be the baby’s ability to use its hand or a fork instead of its mother’s bosom, or the ability of a small bird to catch its own prey instead of eating regurgitated food from a parent Creatures further up the food chain have long enough life spans that they can learn to hunt and forage after they are born Smaller organisms such as snakes and insects somehow already know how to hunt and forage when they are born

In the case of the higher mammals, simply being able to eat regurgitated food is not enough for self-refueling They have to grow

up and learn to capture their own food, or sources of energy Plants just grow Their self-refueling comes based on the quality of the soil and their ability to bend toward sunlight

When we get to humans, in foraging societies, it is not until a male child is about 13 or so that they are taken on hunts Before this age, male and female children are taught to forage and look for plants and nuts that can be used as sources of food also Any organism that cannot capture its own sources of energy dies It is that plain and simple the first feature we look for is self-refueling

If we look at economies, any city or region that cannot produce enough to sustain itself dwindles away Businesses move, people move, etc An economy has to find enough work for its citizens in order to

stay intact In terms of a growing adult, this is relevant as adults have to

go work in the economy to earn money to buy food Depending on how many people find enough work, or create viable businesses for other people to work in, determines whether a city or region gains or looses people

Self-propagating

Live Make babies Repeat

Beyond simply feeding oneself, all living systems have the capacity to reproduce Just about any adult can feed themselves (self-refueling) If they have game (men) or feminine wiles (women) they can create partial copies of themselves, babies, which is what self-propagation is It just so happens that when the majority of organisms are able to fend for themselves and capture their own sources of food, they also become sexually mature enough to reproduce

The feature self-propagating is not necessary for any individual to be considered a living system Self-propagation is necessary for the survival

of the species meday, car companies may even create cars that go get gas when they need it, or are able to plug themselves in at night This would be self-refueling However, it would be a whole different thing for those car companies to be able to create cars that make other cars Cars do not self-propagate Apparently, though, scientists are working

on nano-scale machines that can copy themselves

Similar to our discussion about building networks, the rate at which a population reproduces has to be faster than the rate at which its elders die off If a population falls below this level, then that population will gradually decline If a population stays above that level, then it’ll grow

Habitat preservation

The next feature is habitat preservation Similar to self-propagation, habitat preservation is a feature of the population or community, not anyone individual We hear about the balance of nature all the time We watch the discovery channel and learn that most predators hunt only enough to feed themselves The more food there is the more predators are born Later, as more predators are born because of the bounty, the population of prey drastically decreases After this decrease, the population of predators decreases And the cycle starts back over again This just does not happen with wolves and deer, for example It also happens between deer and their food sources for deer we both have predators and food keeping their numbers in check naturally

Apparently with no guiding hand, nature has complex feedback loops that regulate its environments

But when humans are in the picture, things get disrupted winning author Jared Diamond wrote the book COLLAPSE showing the narrow escape that some human communities have had from ecological disaster This is not some big environmentalist rant This was more an anthropological study in how humans reshape their environment with such speed that the complex interconnections in nature can’t adapt fast enough, and crumble

a feature of the long-term viability of a group of organisms, living systems, is habitat preservation the feature, habitat preservation applies mostly to human communities We can look to the environmental consequences of over-farming and over-fishing and a host of other things Even more recently, we can look to the collapse of ‘high finance’

on Wall Street as a semi-unintentional disregard for the economic environment in which bad mortgages were loaned, bought, sold and leveraged

In either or both of these situations it is hard to say whether there was

an inability or an unwillingness to seriously consider the consequences

of our actions Either way, you can bet your bottom dollar with the perfect storm of ecological degradation and financial collapse that many more people are starting to and will start factoring in the consequences

of their actions, on personal, local and larger levels, into their decisions

A little less gloomy, and closer to networking, we can see the common wisdom of not burning bridges Avoiding doing things that offend people and sever working relationships is a habitat-preserving behavior

Development

I said there were three or four properties of a system that defined it as

a living system Up until now, each of these three previous features is pretty much embedded in any (non-suicidal) individual organism’s behavior Now we are going to look at the changes and varieties inside

a population

When we talk of international economics, we have developed, developing and undeveloped economies We call those economies developed that exhibit a variety of industries creating goods and services for both export and local consumption Those economies that

are primarily geared toward export we call ‘plantation’ economies, whether they export wheat, oil or Nikes

Development is different from growth Growth talks about the sheer number of nodes in a network, people in a population, people employed in a business, etc Development talks about the variety of activities that those people, nodes, perform When we talk of development in a population of say birds, we are talking on an evolutionary scale We talk of the differentiations of birds between one another due to distance, food sources, climate and so on As populations of organisms separate from each other, they adapt to their environment as members of that population that fare best in that environment tend to reproduce more

Biologists have observed populations of squirrels develop different color patterns in four years as one population of squirrels split into two along a newly paved road Recently biologists have also found that ‘big game’ is on average getting smaller as all the big game with the genes for big sizes are poached and hunted far more than others, therefore reproducing less Part of this stems from the thrill of ‘big game’ and part of the reason stems from the laws that lean hunters toward hunting bigger game so as not to kill youngsters what we are seeing here, over generations and not just life-span is the development that I am talking about

This kind of ecological development, over time, is a little different than the development of specialization that we see in an economy Even within a business there must be development, specialization, as not everyone can do accounting and neglect sales within a business it is necessary to have a diversity of skills The same thing happens in an economy, you need farmers, canners, shippers and grocers to distribute food, so you need development of multiple skills to support the population

Looking around

networks can have some features as living systems They can refuel, as when they get at least as many links as they lose You could say they self-propagate by being able to spawn other networks They also contain differentiation or development, as when broad social networks form organizations, and within businesses project groups etc

self-We could even say that there’s some habitat preservation in the form of strategic thinking Okay, that last one was stretching it a little I just wanted to emphasize the point that networks can behave in some ways

like living systems And if you create a network for pleasure, passion or profit, in the long run it would be healthy to consider not just the formation of the group, but also the life-cycle and longevity of it

far we have talked about the basic features of a network Now, if you put this book down and go read books on network theory and social network analysis, you will not be a fish out of water However, some of the books on network thinking talk about the formal properties of networks, but do not translate that into how they actually operate Because they are looking at multiple kinds of networks, it is hard for them to identify what happens in and around a network that makes it work like that Another bias away from being practical for us is that they are mostly academic researchers The academy tends to teach about things, not how to do them do not fault the networking guru’ authors, they do not have a readily available body of literature that translates network theory into specific activities These network gurus are putting their stuff together like a collage And, they are not doing too badly at it either

As these network theorists look at the formal mathematical properties

of networks, they make a lot of assumptions about what has to exist for these networks to be made For instance, they study the interconnections of train stations and airports and traffic, but they do not talk about having to build the trains, rail, planes and airports Let’s say we want to build a high-speed rail network, we must consider building the infrastructure underneath it before we can build the network itself But to do that, we first have to understand what infrastructure is in general, and what infrastructure is in human networks

After we understand what infrastructure is, we can start to build our own infrastructure, or the backbone of our networks The central question to understand and point to infrastructure is: what must exist for this to exist Since not many other books on networking talk about

it, we will have to answer this question for ourselves (the question in the previous paragraph, you have ADD) We will talk about what needs to exist for networks to exist and work: infrastructure, platforms and protocols Later in this book, you will see how this foray into the deeper levels of networks can be very practical, but for now, we will just get our feet muddy