This site, The Landing, is a bit like a building. The more people that enter that building, the more valuable it becomes. The real value and substance of the site is not the building itself but what goes on and what can go on inside it.

If it doesn't provide useful rooms and other spaces that fit the needs of the people within, or if the people inside cannot find the rooms they are looking for, then it needs to be improved - better signposts, easier halls, stairways and elevators, bigger doors, different room layouts. This matters and it's certainly a big part of what influences behaviour: we shape our dwellings and afterwards our dwellings shape our lives, as Churchill put it. However, like nearly all social technologies, the Landing is a soft technology, where many of the structures are not created by architects and designers but by the inhabitants of the space. Far more than in almost any physical building, it is the people, the stuff they share and the ways they share it that make it what it is. They are the ones that decide conventions, rules, methods, procedures, interlinked tools and so on that overlay on the basic edifice to turn it into whatever they need it or want it to be. 

Soft technologies are functionally incomplete. They are needy, by definition lacking every necessary part of the technological assembly that makes them useful. They can become many different technologies by aggregation or integration with other technologies, including not only physical/software tools but also and more significantly methods, norms, processes and patterns that are entirely embodied in human minds.

Hard technologies are those that are more complete, less needy. The more they do what they do without the need to aggregate them with different technologies, the harder they become. All technologies, soft or hard, will play some part in bigger systems and almost all if not all will rely on those systems for not only meaning but also their existence and continued functionality - for example, power, maintenance or, in the case of non-corporeal technologies like laws, pedagogies and management processes, embodiment. However, harder technologies play far more limited, fixed roles in those systems than softer ones. A factory tooled to produce milk bottles probably does that really well, consistently, and fast but, without significant retooling and reorganisation, is not going to produce glass ornaments or thermometers. A metal tube and  furnace need the methods and processes employed by the glass blower to turn raw materials into anything at all but, because there are few limits to those methods and processes and those can be adjusted and adapted almost continuously, can be used in many different ways to create many different things. The needier a technology, the more ways there are to fulfil those needs and consequently the more creative and rich the potential outcomes may be.

A microchip is a very needy technology. Assembled with others, it can become still needier: a computer, for example, is the very personification of neediness, doing nothing and being nothing until we add software to make it be almost anything we want it to be - the universal machine. Conversely, in a watch or a cash register or an automated call answering system it becomes part of something more complete, that does what it does and nothing more - it needs nothing and does what it does: the personification of hardness.

Although automation is a typical feature of harder technologies, it depends entirely on what is being automated and how it is done. Henry Ford's classic production line turned out a lot of similar things, all of them black: it was archetypally hard, a system needing little else intrinsic to the system to make it complete. Automation largely replaced the need for technologies needing skill and decision making to make them complete.  Email, on the other hand, an archetypal soft technology, actually gained softness from automation of (for instance) MIME handling of rich-media enclosures. What was the preserve of technically savvy nerds with a firm grasp of uuencoding tools became open to all with standard for rich media handling that automated a formally manual (and very soft) process. This was possible because automation was aggregated with the existing technology rather than replacing it. The original technology lost absolutely none of its initial softness in the process but instead gained new potential for different ways of being used - photo journals, audio broadcasts, rich scheduling tools and so on. Neediness and automation are not mutually exclusive when that automation augments but does not replace softer processes. Such automation adds new affordances without taking any existing affordances away.

Twitter is a nice example of an incredibly soft social technology that has become yet softer through automation. Twitter is soft because it is can be many different things: it is very malleable, very assemblable with other technologies, very evolvable  and very connectable (both in and out). A big part of what makes it brilliant is that it does one small trick, like a stick or a screwdriver or a wheel and, like those technologies, it needs other technologies, soft or hard, to make it complete. Twitter's evolution demonstrates well how soft technologies are functionally needy.  For instance, hashtags to classify subject matter into sets, and the user of @ symbols to refer to people in nets were not part of its original design. They started as soft technologies - conventions used by tweeters to turn it into a more useful technology for their particular needs, adding new functionality by inventing processes and methods that were aggregated by them with the tool itself. To begin with they were very prone to error and using them was a manual and not altogether trivial process. What happened next is really interesting - the makers of Twitter hardened these technologies and made them function within the Twitter system, and to function well, with efficiency and freedom from error - classic hallmarks of a hard technology. But, far from making Twitter more brittle or harder, this automation of soft technologies actually softened it further. It became softer because Twitter was adding to the assembly, not replacing any part of it, and these additions opened up their own new and interesting adjacent possibilities (mining social nets, recommending and exploring tags, for example). Crucially, the parts that were hardened took absolutely nothing away from what it could do previously: users of Twitter could completely ignore the new functionality if they wished, without suffering at all. 

So, back to the Landing. The Landing is simple toolset with a set of affordances, a needy technology that by itself does almost nothing apart from letting people share, network and communicate. By itself, it is hopeless for almost anything more complex than that, but those capacities make it capable of being a part of a literally infinite possible variety of harder and softer technologies. Only in assembly with social, managerial, pedagogical and other processes does it become closer to or, if that's what people want, further from completeness. And we, its architects, can help soften the system further by adding new tools that augment but do not replace the things it already does, thereby making it needier still, increasing its functional incompleteness by adding new incomplete functions.

It's a funny goal: to intentionally build systems that, as they grow in size and complexity, lack more and more. Systems that actually become less complete the more complete we try to make them. It reminds me a little of fractal figures which, as we zoom in to look at them in greater detail, turn out to be infinitely empty as well as infinitely full.