An Introduction to Applied Linguistics
Linguistics is the scientific study of language and mainly concerns itself with describing and understanding language. It was started by the Indian scholar Pānini in the sixth century BC and formal linguistics developed in Ancient Greece and China sometime in the fourth century B.C. The many areas of study in linguistics tackle everything from meaning, semantics, and stylistics, to theories proposed by Norm Chomsky on competence and performance (individual capacity for language versus its use in different contexts and groups).
Applied linguistics take those lessons and apply them to real-world problems, from language education to translation, and further to natural language processing. Applied linguistics is focused on solving and addressing real-world issues. My focus is more toward the applied linguistics discipline so that I might address the real-world problems in financial services. “In fact, we would argue that contemporary applied linguistics is not so much a field as a way of exploring; it’s a process, a ‘mode of enquiry’ for working with language-related problems and needs.”1 There are technical areas within Financial Services, most notably Knowledge Graphs and Ontologies, that relate more closely to applied linguistics, and for individuals involved in those areas, you will see some common themes. Indeed, the position here is that those disciplines and efforts can be further strengthened and focused through applied linguistics.
I will start with five foundational statements I will rely on as I look at financial services through an applied linguistics lens:
1. Communities of Practice: These are groups defined by their shared culture, functions, and processes that result in a distinct and unique language unto themselves.
2. Language is constantly evolving, and is dependent upon change and diversification.
3. There is no right or wrong language.
4. Language (and therefore data that represents language) always exists in a multitude of forms.
5. Language is a social construct, as opposed to something that can be legislated.
The theme of right or wrong language is the one around which the other four points revolve. The social construct refers to how speakers are blinded by their own view.
We may as individuals be rather fond of our own dialect. This should not make us think, though, that it is actually any better than any other dialect. Dialects are not good or bad, nice or nasty, right or wrong—they are just different from one another, and it is the mark of a civilized society that it tolerates different dialects just as it tolerates different races, religions and sexes.2
I will also refer more to Charles Hockett’s Design Features, especially around interchangeability, the duplication principle, and arbitrariness in regards to these five points.
Surrounding this, and directly applicable to the purpose of this book you are reading, is
[B]elieving that governments and academies can ring-fence a language from outside influence is as naïve as believing that everyone outside the borders of Italy can be prevented from eating pizza or that everyone outside the borders of China can be forced to celebrate the new year without fireworks.3
Further, looking from the other direction, it is a myth that each nation has only one language.4 This myth is a recent one born out of using language as policy to drive centralization and consolidation of power and control.5 In recent times, there are examples of support at the state level in the United States for the restriction or banning of the use of other languages, with up to 30 states adopting English-only statutes since 1981. The motivations typically involve arguments toward the costs to local governments for providing multilingual access. The effect results in the inability of non-English speakers to participate in everything from local government to social services. Across the Atlantic, Proper English was particularly enforced among the London elite to ensure the ability to identify someone not from the upper class. Mincken, in The American Language, explores this extensively, where the protectors of correct English said everything about American was wrong.6 We can look further back into 19th Century Germany, as well. German was made the dominant language of the Habsburg Empire, and the only path for social and economic advancement was to speak German. This led to even Czech-majority regions to shift to German language use and solidified a formerly haphazard collection of communities. 7 While one might argue this standardized language, it is easy to point out that the Czech language is still in existence, and used broadly in a cultural sense. So while power and control may appear to have been achieved, it is more that the problems that existed were simply hidden, more easily ignored, and left to fester.
It is the myth of the existence of a single language that is one of the more dangerous myths, setting easy traps for otherwise sound motivations and goals to fall into. This theme of the dangers of monolithic thinking about language, and the need to instead advocate a pluralistic approach to language in financial systems (and regulations that govern them) will continue to be a primary one throughout this text. This is not to dispute Chomskyan concepts of a universal grammar that ties languages together, but is advocating that there are multiple languages that exist that need to be tied together, as opposed to the assumed existence of a single monolithic financial services language.
Communities of Practice in Financial Services
“The silo mindset does not appear accidentally nor is it a coincidence that most organizations struggle with interdepartmental turf wars. When we take a deeper look at the root cause of these issues, we find that more often than not silos are the result of a conflicted leadership team.”8
“Information silos and poor risk management have cost global banks billions… Modern systems are powerful enough to integrate functions across a bank, but often their adoption is stopped by managers who want to protect their own turf.”9
Anyone familiar with financial services will likely nod along to these comments about silos in the industry. Silos are the bane and blame for almost any and all ills that befall the industry (and many other industries outside of finance) and are oft described in negative terms. They are an easy to conceptualize target. This is because a silo will typical embody very specific cultures and processes that those outside the silo would very much like to change. They would like to change them because that target culture, associated processes, and practices do not match up with the observing party’s own culture, policies and practices. I posit that silos are not an evil mechanism born of turf wars and self-preservation intent on obfuscating the wrongdoings of banks from the world until they culminate in the next Great Depression. Instead, silos are a natural phenomenon that—while certainly the source of many of the problems and market impacts—does not mean that silos are necessarily bad or should be eliminated.
“The great thing about standards is that there are so many of them!”— most often attributed to Andrew S. Tanenbaum10 (although sometimes incorrectly attributed to Grace Hopper).
“The most damaging phrase in the language is we’ve always done it this way!’”11 (This is Grace Hopper.)
Tanenbaum’s quote is typically said with sarcasm and derision. I would like to say that I believe there being so many standards for the same thing is actually a good thing (no sarcasm). Multiple standards for the same thing typically align with different silos. FIX, as a message protocol tends to be dominant within the Front Office silo while SWIFT messaging (based on ISO 15022 and ISO 20022 standards) is predominant in Back Office and Payments silos. Meanwhile, FpML is the main standard for OTC Derivatives. All functionally are communication protocols for sending information from one firm to another.
The Silo and Communities
The term silo is commonly used to refer to some system (operational, technological, or a combination of both) that is unable to (or perceived unable to) operate with other systems. This can be due to any number of forces, from technology to organizational issues and includes things like a lack of shared goals, tools, or communication pathways.
Silos are seen across industries, but are fairly pronounced within financial services. Whether it is a global operations separated from a U.S. (or other country) domestic-specific operation, or a traditional Fixed Income/Equity/Derivatives trading desk separation, silos are seen as having been created by legacy technology infrastructures, fiefdoms created by power-hoarding individuals, lack of organizational capability, or some other negative catalyst. The silo is blamed for a mentality that occurs when departments or groups do not share information, goals, tools, priorities, and processes with other groups or departments. It is blamed for negative impacts to operations, low morale, and is held up as a marker for the future failure of the business.12 Yet, the mentality can also be the creator or at least the maintainer of the silo, making this an even more challenging problem.
I would instead argue that silos are a natural occurring phenomenon, steeped in culture and shared process, that are necessary to ensure work is done efficiently, deep knowledge is created, and an overall better result continues to be produced. This does not mean that they do not present challenges that need to be addressed and solved. There is credence to the claim of the need to break down walls between silos. But the approach to date has not focused on preserving the integrity of those silos, but instead on the elimination of differences in order to create some sort of operational harmony.
The result, and the prevailing go-to solution of many an enlightened leader or consulting agency, is the breakdown of silo walls. A whole industry has been defined by those silo-busting efforts, across books, change agents and consulting firms. The return on investment (ROI) for many is a combination of eliminating legacy or duplicative technology infrastructure, downsizing of staff (with all the positive sounding pejoratives of doing more with less, strategic realignment of staff, and so on), and gaining of efficiencies by being able to use the same data, and therefore, reduce errors and friction. Experience shows that as common as these silo-busting efforts are, so too are reversal efforts.
For example, in the 1990s, Banker’s Trust looked to merge domestic and global settlement operations to simplify systems and processes. But the effort was never completed because of the amount of differences across areas like tax treatment, corporate actions, accounting, and messaging. Further back, the original global operations were actually created when the company looked to expand in 1984 and the lack of expertise in global markets was highlighted by a number of unforeseen problems.
Ian Mailer Sidebar
Ian Mailer worked at Bankers Trust Company13 “In 1984, with the purchase of WM company in Edinburgh, they planned to take custody in house in Europe from AMRO and Barclays to the Harbour side Operation, but there were numerous issues.
1. BT as a US operator, did not have a multicurrency investment system.
2. no expertise in European corporate actions or indeed Global Markets.
3. launch of the Master Trust service in Europe.
4. they attempted to use the WM accounting system as an instruction system and then manually rekey to globenet. Inefficient and costly.
It was a baptism of fire. Globenet could not cope with the fx currency matches required for settlement, resulting in a considerable loss in their first year. I actually had to go to Wall Street and sit with the exec and explain how double entry accounting on the WM system worked!!”
I have always put this down to a cultural difference and legacy market technicalities. What do I mean by that? The U.S. market was relatively simple (and all the better for it). Raising capital in Europe usually required a rights issue, a concept not practiced in the United States at the time. This was hard to explain to the U.S. operations teams. I had become WMs training officer by the time I moved to BT, so was able to try and train the U.S. corporate action (CA) team as quickly as possible. It wasn’t a matter of how good the staff was, just getting them used to the way the rest of the world worked.
Another example was in the early 2000s as the Bank of New York looked to merge domestic and global settlement operations and technology for the purpose of finding efficiencies and shedding older technology and processes. In the end, the technology just put a wrapper around the domestic system that was supposed to be retired—it was never integrated. This would further complicate integrating with Mellon Bank when the two companies merged a few years later.
In 2009, Barclay’s acquired Lehman’s equity traders, allowing Nomura to purchase the technology infrastructure around the Lehman equity trading system. The idea was that Barclays, primarily a Fixed Income house, could easily integrate equity trading into the more complex world of fixed income execution and order management systems. Among many delays, Barclays would have to build an entire equity subsystem, while the equity traders sat idle, unable to trade, for far longer than originally estimated. I attribute this to a failing in appreciation of community’s differences.
In applied linguistics, a community of practice (CoP) is a version of a speech community. Typically, Speech Communities can be well defined through culture, idioms, and language-specific markers that can be used to define them and make them distinct from other speech communities, even if there is some overlap. CoPs are used as a bounding mechanism for applied linguistics to help define a potential speech community because they share some common culture, processes, and procedures that link them together in practice. The actual documentation of the unique speech factors that differentiate one CoP from another may not initially be known or understood at the onset, but the methodology is meant to enable that analysis from an applied linguistic perspective.
Wegner-Trayner defines a Community of Practice as:
A community of practice is a group of people who share a concern or a passion for something they do, and learn how to do it better as they interact regularly. This definition reflects the fundamentally social nature of human learning. It is very broad. It applies to a street gang, whose members learn how to survive in a hostile world, as well as a group of engineers who learn how to design better devices or a group of civil servants who seek to improve service to citizens.
In all cases, the key elements are: The domain: members are brought together by a learning need they share (whether this shared learning need is explicit or not and whether learning is the motivation for their coming together or a by-product of it) The community: their collective learning becomes a bond among them over time (experienced in various ways and thus not a source of homogeneity) The practice: their interactions produce resources that affect their practice (whether they engage in actual practice together or separately)14
In financial services, we use a number of different terms that dance around the concept of CoPs, and come from different perspectives across technology, operations, organizational structure, and business practices. Domain, Silo, Business Unit, Functional Unit, Firm type, Product type, Asset type, and similar terms all speak to a subset of financial services that is bound by a set of shared processes. These terms all speak to the same type of concept, even if the context in which they are used differ. These shared processes are typically unique, or have unique aspects that differentiate them from each other, as well as a culture that embodies the community.
As described in the prior chapter, firm type is one of these concepts that can bound a Community of Practice (i.e., Sell side versus Buy side). But communities can be overlapping, or even subsets, especially at overlapping intersection points (e.g., Buy side mid office). This complex inter-relationship is expressed in Figure 3.1.
Each of these terms infers what is embodied in a CoP; a community bounded by those shared processes (or practice in some linguistic circles), understanding, and culture. All of these must be in place to properly define a CoP. Belonging to one CoP, such as Equity, does not infer that there is shared understanding, practices, or culture between Buy Side and Sell side, or between Front and Back offices.
The practice is important because it identifies knowledge with something people “do” as part of their culture, profession, or avocations. (As any teacher will attest, knowing without doing seems nearly impossible; whatever learning residue exists rarely sticks.) And, as another key insight, Constant says a practice is not enough to specify where knowledge lives, because disconnected groups may share a practice or even a set of practices, but if they are not in contact (harkening back to the idea of a community as a group of people in communion with each other), the meanings of those practices will not be the same.15
Figure 3.1 Community of practice matrix
What is important here is the distinction raised that two groups that share a practice should not be looked at in the same way—if they are not in direct shared communication or interaction, their practices (and therefore their data, language, and interpretations) will differ. Further, it is important to note that when the existence of a Community of Practice is not identified, this is not a wrong thing. Going back to our main tenants concerning linguistics, remember that there is not a right or wrong language, there is simply the fact that a language (a way of speaking) exists.
So, what? Within applied linguistics, the goal is not just to document the language of a CoP. Within the socio-linguistic field, the goal is to solve a problem that is linguistic-based; namely the problems in communicating across these CoPs effectively, the effective sharing of information required by multiple CoPs (adjacent or overlapping), and resolution of conflict between CoPs. Looking back at the complaints about silos, the language used to describe the issues they present center around those exact topics. Let’s remember the complaints about silos. Silos are a representation of conflicted leadership teams. Silos are artificially created to protect turf. Silos are re-enforced by old technology that doesn’t enable sharing of information. Across silos, processes are not aligned. Further, the culture of a particular silo is typically highlighted as something not aligned with the observing group doing the analyzing, and that it is somehow disruptive, or otherwise undesirable.
I propose, instead, to view silos as Communities of Practice. In doing so, we can view a silo through the lens of a community that shares a different culture and language than others with which it must interact. Instead of looking to bust silos, the effort is to be focused on enabling interoperability and language translation between differing CoPs. As I position the discussion in this way, something becomes very clear; much of the conflict between silos, or silos inside a larger organization, can be described as clashes of language and culture. As I refer back to the use of language for politics and control, another aspect of silo busting is the perception that the undesirable silo is somehow misbehaving, or otherwise not aligned with the analyzing group’s goals. By silo-busting, the view is that power and control can be re-established through the imposition of some form of monolingualism, as opposed to approaching the problem from a multilingual perspective.
Language, Evolution, Change, and Diversification
Professor Mark Liberman (University of Pennsylvania) asks as part of an Introduction to Linguistics course: Here is a puzzle: language change is functionally disadvantageous, in that it hinders communication, and it is also negatively evaluated by socially dominant groups. Nevertheless it is a universal fact of human history.16 When I sit back and think about this, language change does seem counter-intuitive. The existence of the thousands of languages globally that exist introduce friction into communication and make it difficult to interact as human populations.
This language change doesn’t only happen among your large traditional languages like Japanese, German, or English. Populations that speak the same language also will drift apart when separated. “In isolated subpopulations speaking the same language, most changes will not be shared [between subpopulations]. As a result, each subgroup will drift apart linguistically, and eventually will not be able to understand one another.17
In traditional human language, changes are affected by such things as learning (one language expert teaching someone new to the language), contact (communication and interaction between groups that differ in language and dialect), culture (societal concepts and norms, from dress to beliefs), and natural use that results in such things as slang or change of meaning. Further, language continues to diverge toward multilingualism as opposed to monolingualism.
This is all the more true as the focus has shifted from Europe’s allegedly monolingual and monocultural national states to a global view, showing that most people cannot live their everyday lives without making use of several linguistic varieties. It seems that multilingualism by far outweighs monolingualism, measured on a historical and global scale (cf. Ludi 1996), an assumption that becomes even more convincing if we accept that there is no straightforward distinction between multilingualism and multilectalism.18
Simplistically, bringing together communities of different languages and agreeing on an overall governance does not unify those communities into a single language, or overrule the continued existence of the underlying communities or CoPs. Relating this to the financial services community, the larger community can be viewed in the lens of multilateralism; it is more of an alliance of multiple Communities of Practice pursuing a goal of an overall functioning within an interconnected financial system. Being an instance of multilateralism, and not a specifically distinct community on its own, the nuances and differentiated data and speech between CoPs in the alliance would not and do not lose distinction. This is consistent with human language.
[T]he general tendency seems to be for the dialects on both sides of a political border to become more and more dissimilar, that is, to diverge. This tendency is in conformity with the “proximity principle” discussed above. Internally, however, state borders tend to have unifying effects, although the linguistic unification is never absolute (cf. Sapir 1921: 213 fn.). This finding, too, is in keeping with the proximity principle. In short, convergence on the dimension of dialect to Standard language and of dialect to dialect (i.e., linguistic unification within state boundaries) necessarily leads to divergence at the borders. Thus political borders that transgress old dialect continua are turning into new dialect borders.19
Relating the various Communities of Practice within financial services to states bound by political borders, it should follow that even when there is some shared language, dialect borders exist and will define where divergence will continue to grow.
Right Versus Wrong Language
“Languages often have alternative expressions for the same thing (‘car’ and ‘auto’), and a given word can carry different senses (‘river bank’ vs. ‘savings bank’) or function as different parts of speech (‘to steal’—verb; ‘a steal’—noun). Because languages naturally adapt to their situations of use and also reflect the social identities of their speakers, linguistic variation is inevitable and natural.
So what is right and wrong in language, and who decides? Some observers claim that the real issue about linguistic right and wrong is one of deciding who wields power and who doesn’t.”20 This idea of power is one that should be of keen interest to those involved with financial services. Especially post 2008, regulators around the world have begun to flex their muscle to bring greater stability, fairness, and access into the financial system. (I do note the seeming lack of focus on optimization, function, and positive outcome regarding the purpose of the financial system’s existence in the first place, however).
One of the more subtly impactful actions, and one that I believe does not garner enough attention, is the focus on the creation or standardization of a common financial language. In December 2017, the European Commission launched a public consultation that refers to the development of a common financial language (European Commission 2017). The Commission’s Financial Data Standardisation (FDS) project also refers to the lack of a common financial language (European Commission 2016, p. 11). This subject was once again raised at the EU Commission Conference “Preparing supervisory reporting for the digital age,” which was held in Brussels on June 4, 2018 (https://bit.ly/2J5TA0R). And it is also represented in the expansive work that has been conducted on the Global Legal Entity Identifier since 2008 as well as in the efforts of the U.S. Office of Financial Research (OFR), which was formed as a result of the Dodd–Frank Act.
In the context of linguistics, it can be observed that various regulators are deciding what is right and wrong in financial services language through the exertion of their given power. Further, this viewpoint toward requiring a common financial language has parallels to the concepts of language as a problem (Ruiz, R).21
The specific outcomes for students is that they are blamed for failing by implying that they are not smart enough, motivated, or appreciate the educational opportunities the school system gives them (Darder 2011). This is deficit thinking by “blaming the victim” (Ryan 1971). These programs stemming from deficit thinking, show bad results worldwide, and assimilate children while at the same time prevent them from getting a good education (Cummins 2001).22
We can relate students in this examination to the various CoPs and actors within the financial system, and regulators to the powers looking for a monolinguistic financial society. The various CoPs are looked at as deficient because of their individual languages, processes, and cultures. Deficit thinking, and the belief that CoPs intentionally behave badly, and primarily use their individual languages to prevent transparency and understanding, simply reinforce regulators’ collective view for the need to force change in all CoPs to a monolinguistic reporting and operational scheme. There seems to be a view that various CoPs intentionally create their own language in order to obfuscate and prevent outside observation and understanding. And therefore that this language is somehow wrong. Meanwhile, as we identified previously regarding change, linguistics tells us that this view is most likely incorrect and the truth is in line with the simple fact that language evolves and diverges, especially within CoPs, with clear distinction at their borders.
Multitude of Forms
According to Wittgenstein’s use theory of meaning, words are not defined by reference to the objects they designate, nor by the mental representations one might associate with them, but by how they are used. For example, this means there is no need to postulate that there is something called good that exists independently of any good deed.23
Meaning-in-context, on the other hand, is less static and more elusive. The meaning of an utterance requires an understanding of its context, a familiarity with the way the utterance is being exchanged, the intention of the utterance, and the position of the utterance within a ‘language game’ or ‘conversation’. Such a theory of meaning must take into account that the subject is a creative, imaginative agent who extends (or projects) new language practices from prior encounters, and that such meaning is framed by the individual’s social and discourse practices.24
What we can indicate here is that a word may differ in meaning based on use and context, data and language in financial services is also subject to a multitude of forms, based upon the CoP that generates that data or language. The same data (i.e., word) generated by two different CoPs not only may, but most likely will, have different meanings. These meanings necessarily will be dependent upon the context within the CoP; how and why it is being exchanged or generated, and where in the process specific to that CoP it is being used.
For example, I want to identify IBM Common stock. If I use a ticker code, that infers IBM common stock in a certain context— namely on the particular exchange or data platform from which that ticker originates. Meanwhile, use of another code, say a SEDOL, also refers to IBM Common stock, but specifically to the settlement regime or clearing regime in which it exists. It is now devoid of the nuances of exchange or data platforms, and ignorant of other clearing and settlement regimes in which IBM common stock may also exist. Yet, in response, if I require use of an ISIN code, which would uniquely and specifically indicate only IBM common stock, I lose the nuances of either exchange or clearing and settlement regimes. The individual objects may be the same, but at the same time, they are very different. And therefore, their meaning and intentions also differ.
Language as a Social Construct
Language, and data that is created in that language, are created by people in order to facilitate communication. They do not exist naturally, but are instead constructed by the social group, the community, that requires the language to communicate information and accomplish its goals. While this may seem obvious, or a simple aside, the implications are important when speaking of Communities of Practice, and the language that is specific to that CoP. As a social construct, language, and the objects they describe, are bound conceptually by the CoP defining them. Multitude of forms and arbitrariness exist partly because language is a social construct, is defined by and evolves subject to a specific CoP’s influence and use.
In Hockett’s design principles, this relates to cultural transmission, in which language is learned through the social setting. While humans are born with innate language capabilities, the expression of concepts are learned, and are a part of learning. The other implication of this is the language that is used by two different CoPs will differ based on the needs and purpose for which that language is used. This is as much a point in understanding not just the language, but the social context in which it comes from.
Infology comes from the work of Börje Langefors, a Swedish engineer and computer scientist. As part of his work, he formulated an infological equation noted as I = I(D, S, t), which describes the difference between data and information. The mathematical expression captures an observation that I, Information, communicated in an information system is a function (i) of the data D, the semantic background S, and the time interval of the communication, t.25
Infology and Decentralization
Information systems theory has, since its beginning in the early 1960s, been facing a contradiction. One of its main visions was that data in the system had to be available to “everybody” (Langefors 1961, #29, 1963, #37). But it was soon detected that a set of data does not inform everybody (the “infological equation.” Langefors 1966, #1). It had to be concluded that efficiently designed information systems had to be structured as a network of communicating more or less separate subsystems based on local data systems. This insight took a surprisingly long time to gain recognition in the data profession, as well as, for instance, in accounting.
Even when, in the 1980s, small local systems came to be fairly common, this was in many cases due to the emergence of inexpensive micro-computers, rather than to an understanding of the often local character of data. 26
What Langefors indicates with the infological equation, and in the aforementioned conclusions, is that local data has its own character, and should be connected into the larger system in a way that preserves its local nature. However, he notes that this truth about data was not recognized. While some may point to the 1980’s decentralization effects, they were not intentional in the means of acknowledging the local character of data. Indeed, centralization continued and continues today to be a major theme and objective in technology and organizations.
Like Machine Learning (ML) and Artificial Intelligence (AI) in 2020, the work of Business Information Systems was, and continues to be focused on using data to provide useful insights (information) that can be communicated to others. The contradiction here is that information systems, including today’s ML and AI, focus on making sure that everybody can access ALL data, while the point is made that some of that data “does not inform everybody.” In other words, there are data that, based on the time period of the data and the semantic background, does not need to, nor should be, exposed to everybody. This is, of course, structured in the view of computer science, and the mechanisms of technology. But we can relate this to our Communities of Practice easily. The semantic background and the concept of local data are both analogous to a specific CoP. In other words, there is local data specific to a community (the semantic background) that doesn’t necessarily have meaning or utility for everybody.
From a pure information systems perspective, subsystems typically perform very specific tasks. They can be specialized and only relevant for niche processes. Therefore, this means that the local data can conflict with or is not relevant to other systems, either because those other systems aggregate data irrelevant to processes the local system is performing, or the other systems only care about a subset of the data generated by any local subsystem.
This is expressed further:
And, with the maturing of the technology of connecting small computers to form networks, one has begun again to talk about making all data accessible to everybody. We conclude that there is still lacking the understanding that some data are only intelligible to restricted groups of people. This suggests that there is need for case studies, in order to reach and disseminate a more concrete understanding of this aspect.
It is often stated, e.g.[,] by data managers, that the popping up of isolated local systems will lead to chaos. Leaving aside the fact that some amount of chaos may be useful, we point out that keeping isolated such data as are in in any case unintelligible outside a limited context can’t by itself generate chaos. Of course, such data as [sic] have to be used in several locations, but those only, must be subject to integrated management—but this should not be done indiscriminately.27
This viewpoint in infology provides the same conceptual structure and reasoning as CoPs. It recognizes that there are local systems, ostensibly dedicated to restricted groups of people for their use—a direct corollary to a CoP. Further, there is understanding that at many times, data will be unintelligible to other people and systems—just as a different language would be between isolated or otherwise contained CoPs. And as with any information system, especially as we look to ML and AI, there is the counter force of wanting to link everything and everybody and provide universal access to all data. But Infology indicates that local data should not be included indiscriminately, and only after integrated management—that is, translation.
Infology, expressed as information being the intersection of data and knowledge, does seem to start down the path we have taken regarding community of practice-focused attentions. However, coming more from a purely mathematical and technical focus, the underlying problem I am looking to address has never really been solved in infology. The debate over centralized and decentralized databases speaks to this continued friction and lack of resolution. Langefors here pays difference to “some data are only intelligible to restricted groups of people,” but indicates that the area is largely unexplored, and instead the focus continues to be on the technical integration and information availability. The issue with focusing on technical integration and information availability is that the nuance of the S in the infological equation is typically de-emphasized, and there is a presumption that technology, applied correctly, can solve all problems.
Technology forges ahead without understanding that some data is not meant for everyone, primarily because it is unintelligible and irrelevant. Chaos is more likely to spread in these systems by including this data instead of understanding why it should be separate. The fear is that chaos will be caused by continuous and spontaneous creation of local systems— called shadow technology, dark systems, or rogue systems, and that this is a bad thing, as it introduces challenges of management, risk, miscommunication, and data loss. Which are all valid points. But again, I point to the fact that language is always evolving. Many times these systems are simply instantiations of that evolution. It is hard to manage and thus the easier method is prevention instead of understanding and evolution.
Accommodation and Fixing
Accommodation and Fixing are tools within linguistics that are used to facilitate dialogue. Simply, when two speakers from different CoPs, languages, or dialects interact, there is expected to be some level of misunderstanding or lack of understanding. In some cases, the speaker or the listener may accommodate the other, able to process non-native concepts or words by self-interpreting the meaning.
For example, a Queen’s English speaker talking to an American may say “I had to wear my jumper because it was chilly today.” The American, not familiar with the word jumper may infer, through both language context and a clue, such as the person pointing to their sweater as they spoke, that jumper means sweater. Further, they do not call out or otherwise question this. Meanwhile, the English speaker may instead use the word sweater instead of jumper, knowing that it may introduce confusion, and thus, accommodates the American speaker. Finally, the American speaker may reply back and use the word jumper instead of sweater in future speech, therefore accommodating their British friend.
Fixing, on the other hand, would involve initiating a query, acknowledging a break in understanding, and resolving that misunderstanding. This is a feedback and repair mechanism. In our previous example, the American may not have had enough context to interpret jumper. Instead, they would ask outright “what is a jumper?” and even perhaps provide their definition of what they thought a jumper was—either a set of jumper cables for a car, or a person on a ledge. A back and forth may occur, until both speakers understand that jumper and sweater are conceptually the same object.
Communication across language barriers must encompass accommodation and fixing in order to solve problems in language and understanding. Translation does not always capture these nuances either, which impacts the ability for technology to consistently solve for such cross-CoP issues. Understanding a second language natively can typically provide a better understanding than a translation from an interpreter. Each new party in the translation introduces their own biases on what may be emphasized or how it is interpreted without the necessary fixing occurring. Much like the childhood game of telephone, where a message is passed down a chain, without any corrective action, and inevitably ends up a distortion of the original message.
The problem of Fixing from a technological perspective is based in trying to fix data misunderstandings across different CoPs, especially when it is assumed that the language is shared and the same. Technology, unless it is expressly told, will not catch language nuance between CoPs in many cases. That is not to say it is not impossible—but without a specific focus and understanding of CoPs and defining them, this task will be haphazard at best, and trying to go off imperfect clues.
Further, capturing the instances of accommodation, especially where members of certain CoPs are naturally multi-lingual, is a larger challenge. There will be cases where users add in extra information automatically to make deductive leaps in regards to accommodating clients and counterparties from different CoPs. While NLP, ML, and AI can get very good within a specific language to ask questions regarding fixing, accommodation is a next level concern.
Wrap these two issues underneath the prevailing assumption that there is, or should only be, one ‘financial language’ and technology has no guide in performing either of these tasks. Technology solutions will gravitate toward a bias for whichever CoP it is most used by, and re-enforce the conflicts and misunderstandings that exist between CoPs when that technology interacts with those that do not share its bias.
1 Hall, C.J., P.H. Smith, and R. Wicaksono. 2011. Mapping Applied Linguistic, 19. New York: Routledge.
2 Trudgill, P. 1994. Dialects. Florence, KY: Ebooks Online Routledge.
3 Hall, C.J., P.H. Smith, and R. Wicaksono. 2011. Mapping Applied Linguistic, 12. New York: Routledge.
4 Hall, C.J., P.H. Smith, and R. Wicaksono. 2011. Mapping Applied Linguistic, 13. New York: Routledge.
5 Hall, C.J., P.H. Smith, and R. Wicaksono. 2011. Mapping Applied Linguistic, 24. New York: Routledge.
6 Mincken, H.L. 1919. The American Language. Knopf.
7 Diglossia and Power: Language Policies and Practices in 19th Century Habsburg Empire, edited by Rosita Rindler Schjerve.
8 Gleeson, B., and M. Rozo. 2013. “The Silo Mentality: How to Break Down the Barriers.” Forbes, October 2. https://forbes.com/sites/brentgleeson/2013/10/02/the-silo-mentality-how-to-break-down-the-barriers/#554ede888c7e (accessed January 4, 2021).
9 Writes Gillian Tett, U.S. managing editor and columnist at the Financial Times (and author of The Silo Effect), Groenfeldt, T. 2015. “Silos can be Costly in Banks.” Forbes, December 28, 2015. https://forbes.com/sites/tomgroenfeldt/2015/12/28/silos-can-be-costly-in-banks/#700ee642356f (accessed January 4, 2021).
10 Most often is, of course, a lazy way to avoid controversy, but is in his book Computer Networks (1981), p. 168.
11 Hopper, R.A.G.M. March 9, 1987. in an Interview in Information Week 52.
12 Roughly interpreted from the Business Dictionary
13 Ian, M. May 15, 2020. E-mail message to author.
14 Wegner-Trayner, E. and B. 2020. “Introduction to Communities of Practice: A Brief Overview of the Concept and It’s Uses.” Wegner-traynor.com: A Partnership, https://wenger-trayner.com/introduction-to-communities-of-practice/ (accessed January 4, 2021).
15 Hoadley, C. 2012. Chapter 12 “What is a Community of Practice and How Can We Support It?” In Theoretical Foundations of Learning Environments, ed. Land, S. and D. Jonassen. https://books.google.com/books?id=FJOoAgAAQBAJ&lr= (accessed January 4, 2021).
16 Liberman, M. n.d. “Linguistics 001 Lecture 22 Language Change.” Introduction to Linguistics Syllabus. https://ling.upenn.edu/courses/Fall_2003/ling001/language_change.html (accessed January 4, 2021).
17 Liberman, M. n.d. “Linguistics 001 Lecture 22 Language Change.” Introduction to Linguistics Syllabus. https://ling.upenn.edu/courses/Fall_2003/ling001/language_change.html (accessed January 4, 2021).
18 Kuhl, K.S. Hoder, and K. Barunmuller, eds. Stability and Divergence in Language Contact: Factors and Mechanisms. Studies in language variation, volume 16. John Benjamins Publishing Company. https://books.google.com/books?id=CKtgBQAAQBAJ&printsec=frontcover#v=onepage&q&f=false (accessed January 2021).
19 Hinskens, F., J.L. Kallen, and J. Taeldeman. n.d. “Merging and Drifting Apart. Convergence and Divergence of Dialects Across Political Borders, International Journal of the Sociology of Language.” no. 145 https://degruyter.com/view/j/ijsl.2000.issue-145/ijsl.2000.145.1/ijsl.2000.145.1.xml (accessed January 2020).
20 Edward, F. 2020. “What is ‘Correct’ Language?” Linguistics Society of America. https://linguisticsociety.org/resource/what-correct-language (accessed February 2020).
21 Ruíz, R. 1984. “Orientations in Language Planning.” NABE Journal 8, pp. 15–34. Doi:10.1080/08855072.1984.10668464
22 Harrison, G. 2007. “Language as a Problem, a Right or a Resource?: A Study of How Bilingual Practitioners See Language Policy Being Enacted in Social Work.” Journal of Social Work 7. Doi:10.1177/1468017307075990
23 “Ludwig Wittgenstein” Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/entries/wittgenstein/ (accessed January 4, 2021).
24 Brace, E. 2014. “Referring to Wittgenstein’s Later Theory of Meaning; Understanding the Relationship Between the Form, Meaning and Use of Language.” https://theliteracybug.com/meaning-form (accessed May 2020).
25 Langefors, B. 1966. Theoretical Analysis of Information Systems, 197. Lund, Studentlitteratur.
26 Langefors, B. 1995. Essays on Infology, 159. Chartwell-Bratt.
27 Langefors, B. 1995. Essays on Infology, 159. Chartwell-Bratt. found via “Centralized versus Decentralized Information systems: A Historical Flashback.” Hugoson, M.A. Jönköping International Business School, Sweden, https://link.springer.com/content/pdf/10.1007%2F978-3-642-03757-3_11.pdf