Lately, I've been studying organizations in my PhD program and I've discovered some things that could help people to understand how the Bujinkan functions as an entity. Over and over I see complaints of people denouncing "rumors" and wanting Soke to be explicit, such as posting official edicts on the www.Bujinkan.com website or the like. A research note entitled, "Note on How Organizations Can Be Structured" from the Harvard Business School was particularly enlightening, and I shall draw heavily on it for this exposition. Hierarchies rest on three organizing pillars: (1) each person in the organization should have one formal direct superior, (2) groups of people at one level should report to a single person at the next level, (3) decisions flow downward from the top of the triangle to the base. Hierarchies are common in militaries and in many traditional martial arts organizations. I will argue that the Bujinkan is instead an "organic" or "cluster" organization. In clusters, action is taken closest to the source of the problem. Knowledge flows across clusters, rather than from "top down." In organic organizations, persons are "expected to be able to make their own decisions to a large extent, and without the constant need of supervision from above." These blurred lines of authority can be frustrating to people expecting a more rigid hierarchical structure. This is what leads people to complain about the rumor mill and the lack of clarity on whether we can train with certain individuals, as an example. In fact, comfort under one structure may mean discomfort under another. It is enough to drive people away to other organizations that meet their comfort needs. Organic structures are extremely flexible and resilient. In hierarchies, jumping a level causes the entire organization to fall apart. This is why people get court-martialed for ignoring a direct order. In cluster structures, however, information jumps from one cluster to the next. This is why "rumors" play such a large role in the communication of information. Soke is quite aware of how information flows in a cluster organization, which is why he rarely makes edicts publicly, such as regarding whether it is okay to train with certain individuals and the like. By definition, people in clusters are EXPECTED to "be able to make their own decisions to a large extent, and without the constant need of supervision from above." So by making a public edict, Soke would be going against the cluster, and instead would be trying to make it a hierachy. I'll give you an example. A few years ago, Soke had commented that the Internet was bad. Seeing the keyboard jockeys and the fraudulent usage of the Internet for gains, he said that we should avoid the Internet. In reality, however, he knew that many of us were involved with the Internet. Well, during that time, Soke made a very grave announcement during training. He said that this needs to be told to everyone. Later in the session, Soke and Noguchi came over to me (this was during the time when I was the designated translator) and repeated to me, "This needs to be told to everyone as quickly as possible." With that glint in his eye, he knew what I would do...and I did it. That evening I sent out a message to people that I knew would pass on the information to others. The clusters had been activated! Within a day, everyone had heard "something." The specifics were clarified over time as more and more people were able to verify from a reliable source. Later, I heard rumors that some people thought it had been "inappropriate" for me to send out that information via the Internet given that Soke had admonished us regarding the Internet. I stood by my choice and took the flak. I knew I was right. I was doing what I believe Soke had wanted me to do. At a later date, Soke commented how the Internet is like a web that ties us all together. That web can be used for good and bad. Ultimately, the implications fall onto the shoulders of those expected to "make their own decisions to a large extent, and without the constant need of supervision from above." I believe Shidoshi and Judan+ must live by these words. I hope this has been helpful in explaining some of the reasons why the Bujinkan works as it does. Whenever you feel frustration with rumors or with the lack of official statements directly from Soke (rather than through a messenger), just remember that the frustration is coming from your desire to see "hierarchical" communication from an organization with an "organic" or "cluster" structure. All the best, -ben PS: This post was originally published on Kutaki no Mura in a thread entitled "Hierarchies vs Clusters."
Wow...that just answered a lot of questions about a NUMBER of organizations I've been having issues with. Coming from a military background, I assumed that these organizations were run like a hierarchy, as that is the basis for which I know and am comfortable with. My perceived "lack of communication" and "dissipation of cohesion" frustrations regarding non-militaristic organizations are now less justified, as I am now a bit more aware of yet another model for the movement of an organization. Thanks! Very informative and well-written post. May you achieve Satori
god i'm never gonna get home... good thread ben...very good indeed. too much spinning round my head and an 50 mile drive impending so i'll post some ideas/questions in the morning. But just to get you thinking..the organic nature of cluster organisations seems to be related to the mathematical concept of self-similar organisation (popularised by fractal geometry) i'm no expert here (hehehe see my comment on your other thread) but it throws up some very interesting questions related to my own area of research. have you considered that a hierachical organisation and a cluster organisation could actually be the same, but just viewed from a different perspective? Under certain conditions an organisation (or organism ) can be viewed from the perspective of a hierarchy and under a different set of observational parameters it can be viewed as a cluster? enough for today...my 'real' life is screaming for some attention
Good post Ben Pretty much how i understand the Bujinkan working. When i need to ask questions i ask my Shidoshi and thats always been good enough for me.
I suppose that a bonus of having Normski as my teacher then, at least i get the information "fresh" and no just heresay or rumours!!
ben, what organisations did your source use an example of clusters? (just curious ) the ideas are certainly valid and an interesting observation is that taijutsu is a route back to a more natural state of being...as such the Bujinkan has spontaneously developed into a system of organisation that closely resembles the way biological sytems evolve to co-ordinate their operation and manage their data flow. I previously introduced the idea of 'self-similar organisation' because it is another way of describing the same organisational paradigm (or way of thinking about things). In essence you can look at a satelite picture of a coast-line and you will see a clear shape, increase the magnification and you will see a smaller stretch of coast-line, but the shape/patterns you see will resemble the original image, and so on. Compare this to the idea of a cluster and each individual cluster has a particular shape or form, yet if you sample a large number of clusters, they are all unique, but a high degree of common features can be found. These inter-related patterns of order have been found to exist all through the natural world. Where i came into contact with this idea, was, funnily enough, in relation to the internet during my under-grad course in electronic engineering. It is one method that the data-flow through the internet can be modelled. It was a really boring part of the course, so i didn't get a particularly deep knowledge of the details and the 'fractal modelling' part was the only bit that caught my attention and wasn't the primary focus My other previous comment about hierarchies/clusters being just different descriptions of the same phenomena; Their is a clear hierarchy of experience within the Bujinkan (from an outsiders perspective) which can be very broadly divided into three groups (excluding soke). At the bottom you have kyu grade...at this level people are in effect 'being conditioned for future training'. Ideas and concepts are introduced at this level and the individual works to internalise core principles and stratergies of action In the middle you have the shodan-godan group, these people have shown that they have grasped the essence of the core principles and are now begining to develop their understanding more completely. At the highest level, you have the godan-judan group. These people have reached a sufficient level of expereince to be viewed as valued sources of information and are the people the lower members of the organisation turn to for clarity and support in the absence of communication from the source. This natural hierarchy then forms itself into clusters and spreads geographically, with the formation of dojo's as the smallest cluster element which will certainly have people from the middle group overseeing the development of those in the lowest group (or youngest group, to express the idea in a less elitist fashion), but not neccesarily containing members of the oldest group (or more senior group, to express the idea in a more diplomatic fashion) Dojo clusters then form into area groups and national groups and as such, at some point contact with the senior grades is made. Thus the natural structure of the organisation is both hierachical in the sense of experience and horizontal in the sense of transmission of information. Bouk, <Removed. Keep comments confined to Bujinkan, we are not going to stray off topic here>
Ben, would it be fair to say that the model you describe has been created to increase the efficiency and speed of information flow towards the decision making source rather than facilitating information flow downwards towards the bottom end of the orgainsation where it is least needed [in terms of decision making]?
from what ben has posted about the model, i understood it to mean that information is better able to flow in both directions. namely that those on the outskirts of a group receive information from the hub just as readily as those near the centre receive information from the edges. With that in mind, it could be argued that in a hierachical structure, information flow is generally restricted from flowing 'up the triangle' to the leaders...and that such a structure would breed the sort of ethos where the leaders were more concerned with their information being received and less inclined to be receptive to feedback 'from the troops' whereas the primary benefit of a cluster would be that the 'leaders' are better disposed to taking information on board from the clusters and that better communication can be acheived in the longer term, once the info has weaved its way through the connecting nodes of the network (slipping into internet language there, but i'm sure you get what i mean ) I'm sure we all know the history of the internet and its design objectives back in the late 60's (it was a US govt funded project to develop a comms system with no central point such that in the event of nuclear attack there was a better chance of comms being maintained, information being able to be routed through a large number of alternate pathways)...thus the cluster idea (which is also a very good analogy of the 'net) allows a greater chance for information to get where it's needed, as it does not rely upon a 'chain of command'. The implication for Bujinkan? well it is clear that dojo-x in a country far from Japan, now has the chance to receive the same info from Japan from a number of sources and the recipient of that info can compare each version of the message, relate it to their personal understanding of the central ethos of the organisation and thus extract the essence of the message from the different versions. This would be reliant upon the ability of the dojo's to correctly interpret the info and has the benefits of encouraging the local decision making and problem solving that is mentioned in the first post on the thread.
ben, i tried to find the article you refer to online and the institution i'm working out of doesn't have a subscription that lets me access that material (unsurprisingly as it is a Psychology Department as opposed to a business school). As the article is not available for free in the public domain i can see why you have not posted a link to it, or a copy, as that would be an infringement of copyright. I also looked at the original thread on Kutakai. The Kutakai thread is interesting and throws up some questions: here is a link to it for those interested: http://www.kutaki.org/modules/newbb/viewtopic.php?viewmode=flat&order=DESC&topic_id=1126&forum=16 (i assume it is okay to link to a different discussion forum? if not will a MOD please delete this part of my post, ie the bit in italics) I did find this article on google; http://www.ftmastering.com/mmo/mmo04_6.htm which seems to provide an overview of clusters from a business perspective. Before asking my questions, could you have a quick scan of this link ben, to make sure i'm coming at the subject from the right angle. much appreciated.
Unfortunately, you have to purchase the article if your library does not provide access via ProQuest or some other such entity. :cry: Actually, this article is referring to the idea of "geographical clustering" of firms. This theory is addressing the question of why Silicon Valley became the "hub" of biotechnology in the U.S. while Taiwan became the hub of chip manufacturing. Even the final section on "Cluster Management" is referring to the same type of phenomenon within conglomerate firms. I am sorry to say that this is not really what I am discussing. :cry: Consulting companies frequently use this type of structure. At any point, people can be pulled "horizontally" into projects to work for clients. Consultants have multiple "bosses" and are not merely hierarchically defined, though there *ARE* different "levels" within the firm (associate, manager, partner). I recommend not getting bogged down by the word "cluster." Instead, you should consider the main ideas most relevant. In short, the Bujinkan is NOT a hierarchy as would be found in a military organization and many, many martial arts organizations. There is simultaneously horizontal and vertical communication, which come to reinforce each other around "the truth." Go ahead and feel free to ask your questions. I'll do my best to answer them. Please do frame them within the context of the Bujinkan, though. Thanks. -ben
i gathered as much...which is why i asked for clarity...thanks for that. good point...i was just picking 'salient' words out to throw at google to try to find some more info about the model in the absence of the referenced article. just to tie the above link down...with what you learnt from the original article...would you say that the core ideas have been applied to the 'geographical' aspect of tech-firms in the way you are suggesting the core idea's apply to the Bujinkan? my actual queries... your post implies/states that the nature of the Bujinkan differs from more orthodox MA structures in that the hierachy we are used to is not in fact how the Bujinkan is geared to distribute information. In your experience, do you think that such a change/development from hierachical structures is a good thing in all regards? Or could some of the 'problems' that surround the Bujinkan be attributed to this difference in structure? The problems I refer to are the differences in the perceptions of just what is and is not 'Bujinkan'. Examples off the top of my head include the whole SKH problem, where their exists a wealth of published material with 'this man is the western authority' all over them, people training for a given period of time within the Bujinkan, feeling they have 'got it' and then leaving (including Japanese shihan and western dan-grades), the whole 'lineage' debates which fill this forum and others and even the central question of just what constitutes 'ninjutsu' training (ie the fact that the concept of 'ninpo taijutsu' was replaced within Bujinkan by the term 'budo' taijutsu). These are the problems facing someone 'outside the Bujinkan loop'. For people inside the organisation...a similar confusion could arise, not in referance to the above topics, but in matters more directly related to the business of training. A particular student could pick up something from a visit to Japan, and not be understanding it in the way Hatsumi was meaning it (if you get what I mean). They then return to their dojo, and share their 'new' insight. This mis-interpretation of the core message could then spread horizontally to other dojo's and it could be some time before the shihan in Japan realise what is spreading through the organisation is 'off-message', making it harder to correct. To people within the Bujinkan, these problems can be resolved to a certain degree by consulting more experienced practioners, but following the logical implications of the model, there could exist, within a 'cluster' organisation, a number of different 'truths'. Each one based upon many variables, depending upon how and when the information flowed to them. I hope that rambling makes some sense
Yes.... In *ALL* regards, no. But I would say the same thing about Hierarchies in *ALL* regards, as well. Yes, that's precisely why I wrote the article. People who do not understand that the Bujinkan is a cluster organization rather than a hierarchy certainly *DO* get confused, imo. It wasn't too long ago, for example, that we heard comments, such as, i personally can't understand why or how a grade system which does not fulfill this basic criteria needs to exist. It is not serving any purpose, because , as a metric, it is ineffective. These type of "problems" (as you call them) manifest from a lack of understanding of the Bujinkan as a cluster organization, in my opinion. Naturally, thus the article.... Okay. This is clearly where I should spend some time.... A particular student could pick up something from a visit to Japan, and not be understanding it in the way Hatsumi was meaning it (if you get what I mean). They then return to their dojo, and share their 'new' insight. Yes.... This mis-interpretation of the core message could then spread horizontally to other dojo's Yes.... But please recall that this will not happen IF THE OTHERS ARE ALSO KEEPING CURRENT with training in Japan! That's the key!!! it could be some time before the shihan in Japan realise what is spreading through the organisation is 'off-message', making it harder to correct. Again, the Bujinkan is not a hierarchy... So why should it fall to the "shihan in Japan" to "fix" the off message????? This is a key point. A cluster organization with many members going to the source on a regular (and most probably staggered) basis will minimize the "off message" messages. So long as people are "plugged in" to the greater community, through regular sharing and seminars, the "shihan in Japan" will not *NEED* to "fix" any problems. THE PROBLEMS WILL FIX THEMSELVES! It's truly a beautiful system, in my opinion. The folks in San Francisco have a wonderful system of sharing, as an example. Not only are there, (what is it Dale?), at least five Shidoshi in the area, there are seminars from outsiders on an almost monthly basis. The Shidoshi travel to Japan with their students at different times of the year, sometimes more than once a year. The seminar instructors also bring an interesting spin on things as well. In the end, there is a continuity between all of these teachings and all of these teachers. That continuity...that THREAD, so to speak, is the essence of what Hatsumi-sensei is teaching!!! The Bujinkan is truly an amazing organism, imo. I know a lot of people are turned off by not having direct "orders" down along ranks, but I personally do not see any of these so-called "problems" as real problems. They are just a symptom that people need to read my article! LOL! Oh, and train with more people. And then, of course, go to Japan on a regular basis. Wrap them all together and you have the perfect recipe for "success".... :love: Training in the Bujinkan is not like collecting fossils and labeling everything with little pieces of paper. The Bujinkan is like tending a garden, with beautiful discoveries emerging every day! Hope that helps! -ben
Correct. I would only comment additionally that quite a number of non-shidoshi level students in the area, both individually and in small groups, also travel periodically to Japan at various times through the year on their own, without their instructors. When they return, several classes will generally get turned over to them and they'll be stuck as the "duty instructor" for a bit, to get them to try to share some of what they've learned. I've long since lost count of how many times I've been accused of "channeling" Nagato, Noguchi, or even Hatsumi sensei by returning students and shidoshi, or laughingly accused of "astral espionage": I'll be showing something we were working on while they were gone, and be told, "But X was showing us exactly that last week in Japan, and making the same 'key points' about it!!" It comes from this sort of "Brownian motion" among the students and groups in the area on a constant basis. It's also facilitated by the fact that a lot of students "cross-train" regularly with more than one group and instructor (though they'll still retain a single teacher for ranking purposes). And to be honest, some of us who are teaching don't really even keep track of who is "technically" whose student or who should be ranking them. I wouldn't say "It Takes A Village" to take care of a student's progress -- sorry, I really don't want to credit Hillary Clinton for ANYTHING!! -- but here, it seems, it sometimes takes a clan. Student rank promotions are often a result of informal conversation and consensus among several shidoshi, one of whom then volunteers to send in the guy's or gal's money to Japan. Needless to say, all this exposure to Soke and the shihan by all sorts of people across all levels -- local students, local shidoshi, and "imported" seminar instructors -- is also a powerful counter-force to any sort of attempt at "empire-building" or any insinuation by any instructor that he/she has a monopoly on "the way". It also ensures that even people who are not able to get to Japan themselves have constant exposure throughout the year to people who've recently been there, and we all stay very much "in the flow" of what's being taught. It's the next-best thing to being there.
On the whole Ben I enjoyed reading your thoughts on the organisational structure of the Buj. It stands to reason that attempting to describe how a MA group passes information through the structure is an extremely complex task. What I have done is presented an alternative to your suggestion. I am not saying I am right and you are wrong as, ultimately, we will both be right (and wrong for that matter). Since nobody has challenged Xenmaster’s description of how the group is organised I will base my understanding on that model but introduce a slight variance. Essentially management models, whether they are hierarchy based, matrix based, cluster based or a variant of these three, are created in an attempt to describe how information flows through a group towards the decision-making person or body. Some groups require everyone to assimilate some of the info before it is passed upwards while others do not. Some groups require the information to be passed laterally before being passed upwards to the decision making body. Martial arts group are an enigma when it comes to describing how they are formed, organised and structure. I feel it is worth making two distinct descriptions in what type of information is passed though the group in order to better understand the model. These distinctions are global knowledge and local dojo info. Global Knowledge: Knowledge essentially follows a cascading path downwards through the group. Hastumi passes knowledge to his Shihans and down though the chain. In a very basic sense, it could be said that, in terms of knowledge transmission, it could be said that the Buj follows a variant of the Hierarchy model. The main difference is that the direction of “information flow” is reversed. It doesn’t serve any function for the health of the group for the knowledge to continually miss a level both in terms of group survival and in teacher student relationships. In my opinion there are strong links between the health of the group and the strength of the relationship between teacher and student. To add to the model it can be said that lateral organisation channels exist between equivalent levels in the group. In other words, levels learn from each other and pass this knowledge to each other. Although Hatsumi and his senior grades learn something from training with the students, it is clear that the senior members are the ones who determine how and when knowledge is released into the wider network. As Hatsumi is Soke, ultimately, it is at his discretion as to what knowledge is released. Instead of describing important levels as decision makers (as in a business model) in a MA model it could be better to describe them as knowledge providers. Could it be said that, in terms of knowledge transmission, the Buj would be better described as a hierarchy – a Division Design? In this model knowledge processing capacity is expanded. This has a significant positive - more knowledge providers exist to divide up the load. Not everyone needs to see Hatsumi in order to get access to the knowledge. Nor would they need to wait for knowledge to pass though a strict hierarchy. That would be far too time consuming and ineffective. To complicate things even further I would like to define two types of lateral organisation channels – temporary and semi-permanent. Temporary lateral organisation channels can be used to describe the one-off knowledge sharing events that exist within the group such as Dale describes. There are moments when everyone will contribute something or some piece of knowledge to the training group. Semi-permanent lateral organisation channels is a term I would use to describe those events that take place regularly, such as Courses and Tai-Kai’s or even trips to Japan. I would suggest that, although each lateral organisation channel is vital to progress of the organisation and serves to increase the knowledge transmission capacity, they are few are fair between as compared to the bulk of knowledge transmission. In other words, the fundamental process of teaching must logically follow a downward path from Shihan to Shidoshi to Student based around hierarchical divisions. Local Dojo Info: Local dojo information is exactly that; local! This includes all the “in-and-outs” or logistics of running a dojo. It may seem academic to talk about this aspect but it still holds relevance to the organisation model so it deserves a brief comment. I will again make use of Xenmaster’s assumption seeing as they have not been challenged. It goes without saying that the local dojo structure is based on geographical location. This is where the Bujinkan is no different from any other MA organisation. Generally speaking there is always a hierarchical structure to the local dojo based on grades or delegation of responsibility. Some areas may have several dojos under the responsibility of one person. I assume that, in this case, further responsibility is delegated to others to facilitate the running of a single dojo. Again, the description of a division design fits nicely when considering local dojo information. Does Hatsumi need to know the problems of a dojo in the South West of England? Of course not, responsibility (or decision making) for this group is delegated down the chain to the appropriate level (and the delegation to a sub-group down the chain further still). Dojos are run, however, under guidelines set by Hatsumi to be interpreted by the most senior member of the hieratical division. Are the divisions completely autonomous? Possibly not due to this reason. They could well be considered completely independent from each other though. Would a Shidoshi in America make decisions on the internal running of a group in England? In terms of logistics and functionally the local dojo could be considered to be hieratical. Cluster organisations – Effectively cluster organisations are nothing more than a variant on a hierarchy division design. As with the division design the clusters / cells can be completely autonomous in nature without needing significant ties from outside of the group to function. Does the description of cluster design fit with Ben’s description of how the Buj operates? I obviously cannot answer that as I am not a member of the Buj but what I will do is present some of the ideas from the note referenced by Ben [1] seeing as he drew “heavily on it” and then present some questions. Friesen and Mills describe principle of cluster design lying in short lived nature of the information being analysed. This is most likely why groups such as investment banks base their management structure on such a model. Simply put, by the time the information gets passed to the decision making person or body, the information may have changed and may not be reliable. Therefore, decisions must be taken, Perhaps this is an “outsider’s” perspective but I fail to see how this principle could be applied to a MA group. Is the knowledge in such a state of flux? From how Ben and Dale have described the group there does seem to be some correlation between Friesen and Mill’s description of blurred authority lines in favour of, and the type transfers of information in the Buj. It strikes me though that the obvious question is how reliable are some of the lateral organisation channels? The main weakness of the cluster structure are described in [1] as, Over reliance on lateral organisation channels as a method of gaining knowledge essentially means that everyone, at some stage, is forced into becoming a knowledge provider. It stands to reason that not everyone has the skill set (or the “specialisation”) to be able to do this effectively. Some people are good at teaching; some people are bad. The amount of understanding the person has on the Art must effect the quality of information being passed to others. Can a 5th Kyu pass on the fine details of Hatsumi’s message as well as a 5th Dan or alternatively a 5th Dan as well as a 10th Dan? Perhaps the over reliance on these channels by some are the reason why so many rumours and falsehoods start within the group? Core Principles - What doesn’t fit into the model is the transmission of the core principles of our Art. How does the development of Taijitsu fit into a cluster model or divisional model? Well, it probably doesn’t! The question I would put to Buj members is – does the transmission of knowledge get given a higher priority over the development of Taijitsu? It does strike me as strange when people ask about an instructor that answers seem to be based around, Now, that may seem like a perfectly acceptable response for some. After all, if getting the latest Hatsumi knowledge is what you are after then you may need to know how often the instructor trains with Hatsumi! However, wouldn’t a more appropriate response be, Shouldn’t an instructor be rated on how well he can perform the core principles of the Art rather than how much information is contained in his folder or how many stamps are on his passport? It does lead to an interesting question. Where does Hatsumi get all his information? Does all of it stem from his time with Takamatsu or does most of what he teach stem from having a high degree of understanding of the core principles of the Art and developing info from that? I would suggest than the answer is somewhere in between. It stands to reason that if an Art that is not refined or added too it will die out as the subject matter will slowly decrease over time. So it would make sense to conclude that the Art is “added too” by men. If that is the case then how did Hatsumi get to such a position? I would like to introduce an analogy given to my by a friend. Each of us learns to write by putting pencil to paper and drawing individual letters. These letters go onto form words; words go on to form sentences; sentences form paragraphs. Before you know it large texts can be created. The question is: what sets somebody who writes for the local rag apart from somebody who creates a inspired piece of prose? After all, each "knows" the mechanics of writing the English language. At some point each of us starts to develop our own style of writing based on fundamental lessons. Just as a person of considerable developed skill adds to the English language by creating great works, it stands to reason that a person of developed skill can add to the Art of Ninjitsu once he has understood and developed the core principles of the Art. I believe Hatsumi is creating "great works of prose" because he understands how to write letters and words and has developed his skill to such an extent. Leading on even further. Is the thirst for knowledge without the understanding of the core principles of the Art a contradiction or ultimately a fruitless endeavour? If Hatsumi’s position is the Holy Grail of training (he describes himself as zero) then wouldn’t it make sense to focus on the reason why he is able to develop such knowledge? Would it make more sense to focus on developing Taijitsu rather than expanding the Ninjitsu “vocabulary”? Without the understanding of the “language” the words are effectively useless. How can men, at some stage, be expected to develop their own “knowledge” without understanding the core principles? Reference: [1] FRIESEN, G.B. and MILLS, D.Q., “Note on How Organisations Can Be Structured,” Harvard Business School, 9-490-040, Dec. 1989.
