Let’s Make Music Together!

Author: Unknown. Link to original: https://www.taus.net/think-tank/articles/translate-articles/lets-make-music-together (English).
Tags: НУА Submitted by LevchenkoAS 31.03.2017. Public material.

Translations of this material:

into Russian: Сделаем музыку вместе!. 5% translated in draft.
Submitted for translation by LevchenkoAS 31.03.2017


A few years ago, I converted my garage into a music studio to support my musical hobbies and interests. The initial investment I made in time and money has paid off in that I have spent countless hours of creative bliss experimenting with music on my own, but although I enjoy my hobbies immensely, I always come to the same conclusion after each session—music is social; I need other musicians to help me fully experience the phenomenon of creative expression through music. I need a group. I need a band.

Calling a rock band together isn’t too difficult. I already have all the space, instruments, and equipment I need to host “jam sessions”. I can even record and produce the music on my own. Music is universally appealing and many people seem to share the same enthusiasm I have for collaborative music creation. Plus, there is no shortage of people who want to be rock stars.

Working with a rock band, however, is difficult. The reason it is difficult is not merely a matter of “prima donna attitudes”, “personality conflicts” or “people problems”. The main problem is that there is no standard body of knowledge required of amateur rock musicians. Nobody is at the same level of skill or experience and nobody seems to know the same songs. Without standards, jam sessions quickly devolve into chaos. Without some unifying principle to coordinate the effort, the improvisational rock band is reduced to a collection of individuals spontaneously and simultaneously playing fragments of the musical patterns they know whether it fits the song or not. Except for a few serendipitous moments of musical luck, the whole experience is frustrating and disappointing.

Jazz musicians don’t have this problem. Over the years, jazz has developed a catalogue of “standards” comprised of about 300 songs that every jazz musician knows how to play. And each of those songs is derived from a well-known, well-understood musical construct like the 12-bar blues. This not only solves the frustration of the jam session experience, it also provides the basis for truly creative improvisation because it allows musicians to embellish and deviate from known standards within a commonly accepted framework. Ironically, musical standards held in common actually encourage and inspire new ideas, new songs.

Another benefit of standards for jazz music is that jazz musicians enjoy a professional community of fellow musicians capable of making spontaneous, lively music anytime, anyplace, and with anyone. Now that is the kind of freedom of expression in music that I want! Music is not an individual project, music is a group experience!

The need for standards to encourage innovation and collaboration in a community of enthusiasts is not limited, of course, to music. In fact, wherever you have a group of people trying to cooperate on common goals, you will find a need for standards. The language technology community is no exception. At present, for better or worse, the language technology community is like the amateur rock band that just can’t quite pull it all together. Using the analogy of the jam session experience as an example, I submit that this problem can be solved by focusing on two principles: interoperability and open standards.

What is interoperability?

Interoperability is the ability of two separate and distinct systems to interact and operate across physical or logical boundaries. Any time there is an exchange across a boundary, interoperability comes into play. When a guitarist builds a layered effect by connecting two “stomp boxes” together, for example, there is an interoperability scenario in play. Stomp Box A receives and processes a signal from Stomp Box B.

Fortunately for guitarists, the interoperability issue is circumvented from the start because the interfaces to the guitars, amplifiers, stomp boxes, mixers, and other signal processors needed to route analog audio signals all accept the ¼ inch (6.35 mm) TRS phone plug standard invented in 1878 to facilitate manual routing of telephone calls by the telephone switchboard operator. Previous to that, telephones were rented in pairs and could only communicate on a single line with their mate. Did Alexander Graham Bell ever imagine that less than a century later users of his audio signal routing technology would reroute the audio signal from electric guitars to stomp boxes to intentionally add distortion and reverb before sending it back to the amplifier? What noise!

For the sake of clarity, the concept of interoperability should be further categorized as syntactic interoperability and semantic interoperability. Syntactic interoperability means two systems can exchange data because they use a common file format or communication protocol—a common syntax.

Data is moved from point A to point B. Semantic interoperability goes beyond syntactic interoperability with the additional condition that the two systems interpret the data unambiguously—a common meaning. Value X in System A means the same thing as Value X in System B. Semantic interoperability is only possible if there is a common reference used by both systems to disambiguate data values, thus the need for standards.

Why is interoperability important?

One of the pitfalls of interoperability initiatives is to settle for syntactic interoperability when semantic interoperability is required, to fall short of agreeing upon and implementing a standard. For evidence of this, you need to look no further than your own experience of using multiple editors to create web pages, word processing documents, or graphic files over the last 20 years—the input and the output don’t match. To use a familiar phrase, something is lost in translation. All the data needs to be there and it needs to be interpreted and processed as intended or you don’t have true interoperability. On the whole we are getting better at this, but the union of syntactic and semantic interoperability through standard specifications is still one of the main challenges of implementing interoperable systems.

Other industries have made great strides in interoperability over the years. I already mentioned the telecommunications industry. The ability to make an international call without the assistance of an operator is a testament to the advances of interoperability standards and mechanisms deployed in communications technology.

The travel industry has allowed online reservations and ticketing for years through their mainframe system with thousands of suppliers including airlines, hotels, and rental car outlets. Recently, through the work of the Open Travel Association, an open standard based on XML and Web Services has been developed to make the system even more versatile. The international banking industry uses a common standard called SWIFT to transfer money between member banks. Thank goodness they thought about semantic interoperability! Otherwise, I might be paid in potatoes! Or worse, I might owe you some potatoes!

In short, due to the nature of our interconnected world, successful interoperability is paramount to achieve the goals of most of the coordinated systems of our lives. Consider the importance of the European Union’s Interoperability Framework to promote eGoverment Services, FATPOT in the state of Utah to facilitate cross-agency responses to emergencies concerning law enforcement, firefighters, and medical services, or the W3C’s mission to promote interconnectivity on the web.

In our own domain, important work is being done by both formal and informal organizations to promote standards and interoperability for localization. Many useful standards have come from OASIS, including the XLIFF standard and the OAXAL reference standard. Other groups, like Interoperability NOW, have declared through the Interoperability Manifesto the need for a grass roots coalition to promote cooperation across the industry. Even TAUS is stepping up to the plate by taking on interoperability initiatives including quality evaluation metrics, the interoperability dashboard, and the language technology on the web project. Interoperability is the order of the day.

Open Standards

Interoperability can be implemented either looking forward or looking back. Either way, standards are required. In the case of looking back, a de facto standard based on market share of an old technology is hacked together to foster goodwill and usually amounts to basic syntactic interoperability. In the case of looking forward, though, an open standard is required to design for the needs of the future.

Any discussion of open standards seems to start with a debate about the definition of the term. Rather than add to the confusion, though, let me clarify by saying that in my mind open standards means extensible standards—standards that are rigid enough to serve a useful purpose as a guide or framework, but are flexible enough to allow for adaptation, customization, and evolution without the usual trappings of custom, proprietary, and closed formats. Open standards are open precisely because they are visible and accessible without restriction. Like the catalogue of jazz standards, they are a given and thus become transparent to the point of becoming a non-issue. Open standards are the foundation of both innovation and interoperability.

The common concerns about adopting and implementing open standards include the overhead of supporting superfluous features, perceived restriction of creative freedom, dilution of competitive advantage, and divulgence of proprietary information. All of these concerns are rooted in the same belief—opportunity is scarce. This attitude is understandable in light of the fierce competition in the marketplace, but ironically, it is the greatest obstacle to success. Monopolies based on closed, proprietary technologies lead to stagnation and market failure. If that happens, everyone loses.

It is widely held that a thriving marketplace requires healthy competition, but paradoxically, healthy competition requires cooperation based on open standards. Otherwise there is no frame of reference to compare quality or to measure progress. There is plenty of evidence across multiple industries to support the claims that opportunity is plentiful and that open standards will actually increase the size of the marketplace. If that happens, everyone wins.

Open standards imply interoperability because the purpose of the standard is to agree upon a basic format or protocol used to exchange and interpret information. The unencumbered exchange of data is the key to market growth because it is the only scenario that promotes innovation based on true improvements in quality and service. By removing the dependency on proprietary systems, the customer is empowered to choose the best solution based on merits. The focus is on desired outcomes and experiences, not on the limitations of legacy formats, mechanisms, or processes.

Consider the benefit of open standards on the Worldwide Web, for example. Web browsers are the interface to mobile and cloud technologies, stateless computing paradigms built on the open standards of the Worldwide Web Consortium that have opened the door to the information age. Web browsers are merely portals to our data. The true value of the web browser (and, therefore, the competitive advantage) is the ability to make data access ubiquitous. The content is more important than the mechanism. Web browsers compete on their ability to deliver content, not on how to format the content. No web browser developer advertises that their product supports HTML. As an open standard, this is a non-issue, but as the chief mechanism of content delivery over the web, it is the basis of the numerous data access possibilities we are able to entertain today.

Do you remember the browser wars of the early 1990s when HTML was not firmly established as a standard? Do you remember the days of conditional parsing of web pages based on browser manufacturer and competition based on format? Do we really want to go back?

How do we achieve it? A three-pronged approach

By their nature, open standards cannot be developed in isolation. This is both a blessing and a curse. It is a blessing because interoperability inherently requires negotiation between two parties or systems. One side can’t define the requirements without agreement from the other side. It is a curse for the same reason—debate and dialogue take time. Decision by consensus usually deteriorates into inaction.

The Need for Standards Organizations

I suggest a three-pronged approach to balance the need for consensus with the need for action and both of those with the need for context and relevance. The first prong is to work through credible standards organizations to define open standards for the localization industry. The work Andrzej Zydron is doing with OASIS to promote the OAXAL reference standard is a good example of how to do this. If implemented broadly in industry standard tools and systems, the OAXAL standard will serve to integrate localization data into structured DITA content using the xml:tm namespace to provide inline authoring and translation memories. By considering the localization workflow from end-to-end, the OAXAL standard is able to solve systemic interoperability problems at the source and avoids the trappings of proprietary formats by building upon established standards like XML and Unicode. Working through formal channels in this way, Andrzej is able to affect the industry as a whole.

The Need for Grass Roots Activists

The second prong is a cadre of outspoken activists that push things forward. Though informal and loosely organized, this group is no less important to the shaping of interoperability standards in the industry. The work that Interoperability NOW is doing to promote open technical exchanges between competitors is a good example of a practical approach to cooperation across the industry. MemoQfest, organized by Istvan Lengyel, is an excellent forum for this kind of technical exchange.

Individuals can also contribute as change agents. Brian McConnell of Worldwide Lexicon, for example, is defining the interfaces for a standard RESTful API that builds on the original translation web services idea put forward by OASIS in the early 2000s. Individual action is the quickest way to get something done. The work that Brian is doing will certainly percolate throughout the industry and it may well be the foundation of my personal vision of implementing real-time translation on localization streams akin to audio signal routing and processing.

The Role of TAUS

The third prong is a think-tank organization that can advise the industry on relevant issues and trends. With the announcement that TAUS will operate as an interoperability watchdog, TAUS is poised to be the conductor and coordinator of our collective interoperability efforts. TAUS is not unlike the judiciary branch that balances the roles of the executive and legislative branches in our three-pronged approach. (See, I did learn something in civics class!) Important initiatives to help judge the merits of our work include quality evaluation metrics initiatives, the interoperability dashboard, and the language technology on the web project. By acting as an intermediary between the formal and informal approaches, TAUS can strike the balance between consensus and action and provide the necessary context and insight to inspire and give purpose to our cause.

Let’s Make Music Together!

I have a room full of technological wonders and so do you. I have a head full of knowledge and so do you. I have a life full of experience and so do you. I want to make music and so do you. With a renewed focus on standards and interoperability, I think we can connect it all together. Bring your gear and let’s jam!