Not so today. By publishing timetables on a web site derived directly from a scheduling database, the railroad can update its schedule as often as desired, perhaps even daily, and travelers can be assured of always finding the most up to date information by checking the appropriate web site. In many ways, this represents a vast improvement for myself as a train passenger, as well as for the railroad company.

But a wealth of information exists in that 1928 timetable that in all likelihood will not exist in the electronic version 75 years from now. The electronic version will not tell me that the train once stopped at 43rd Street, or that it took 27 minutes to get there from the Harrison Street Station. The electronic version will not allow us to figure out that this is the train and station that someone is referring to in their autobiography (or that their account of taking it must be fraudulent, since we know that it took 27 minutes, and they claim to have made the journey in 12). And the electronic version will not have someone’s handwriting on it, showing the address to which they were traveling on that day in 1928.

Consider the ephemera that our culture seems to love and loathe the most simultaneously: advertising. Even those of us who complain most loudly about the ubiquity of marketing messages appearing on every available surface often enjoy and admire the advertising of the past as interesting, funny, or warmly nostalgic and evocative. (I myself, a goatee wearer, have a Gillette advertisement proudly hanging in my bathroom, firmly admonishing “Jobs often lost… because a man neglects shaving!”)

These are more than mere kitsch. Ephemera such as these immediately convey a wealth of cultural information beyond their simple content. Ephemera displayed in museums, used on the stage, or employed in film sets the contexts in which to consider a subject in ways a suit of clothes, style of architecture, or period sofa could never accomplish.

It could be argued that the artistic and cultural explosion that has occurred in the 550 years since that first broadside was printed directly relates to the availability of such common reference points and contextual triggers that these ephemera provide. Like science, creativity is built on the foundation of what has come before, and is empowered by our shared understanding of that heritage.

The possible peril of hyperphemera is that the high availability of these shared cultural reference points will begin to disappear. Even if a high percentage of the web equivalents of ephemera remain stored and accessible, they will lack the persistence of ephemera.

Striking a balance
It used to be that businesses and governments were reluctant to eliminate printed communication because of lingering doubt about the staying power of the Internet. While the web as we know it is a technology that will surely pass, the ubiquity of the network seems assured in the absence of a global catastrophe or impossible-to-predict social or technological revolution.

In other words, within a relatively short period of time, there won’t be any technological barriers to the hyperphemeral approach to disseminate most information. The question becomes, then: what factors can we use to evaluate the impact of moving to a strictly hyperphemeral model in a given instance? I have a few preliminary suggestions:

Speed of information change

Without a doubt, hyperphemera enables much more efficient dissemination of information in many cases. For example, while some newspapers still publish the closing prices of selected securities in their daily editions, the actual second-by-second fluctuations of these prices make any sort of more permanent published record cumbersome and impractical (as much as some of us may love the idea of the clickety-clack of a ticker-tape machine as background noise).

In contrast, the railroad schedule should be much more stable and does not benefit from being reduced exclusively to hyperphemeral information. Indeed, the customers of the railroad rely heavily on the fact that in the ordinary course of events, the schedule does not change very much.

Adaptability of information to statistical analysis

Again, using stock prices as an example, hyperphemera holds no disadvantage because the historical data is most easily represented statistically, showing information across periods of time within charted snapshots rather than in blocks of data. Anyone who had an ephemeral record of the information would probably only use it to perform these types of statistical analyses anyway.

But a catalog and price list, despite being overwhelmingly dominated by numbers, is not statistical. Some of its usefulness can be replicated in hyperphemera, such as the ability to browse through “pages” of products and prices. But much of its lasting usefulness might be largely lost, such as the ability to easily find a product that you remember someone once carried, the opportunity to see at a glance just how much the price of hard drives has gone down over the last ten years, or the ability of your customer to easily keep a handwritten annotation (‘KF9030 same as KF8990’) on the master reference, rather than on an easily lost piece of paper.

Role of serendipity in user tasks

Self-contained, time-limited tasks work well in a hyperphemeral model. It’s relatively unlikely that someone will benefit from serendipity when filling out their shipping information or applying for a credit card. Similarly, if choices truly are limited in a shopping cart model, hyperphemera holds few disadvantages.

But a lot of information isn’t used like that. Using the railroad again, it’s very useful for me to use the web to find out that my train is at 10:25, until I’m at a meeting and realize that I’m going to miss that train. In such a case, the ephemeral reference is superior, because it affords easily accessible contingency options. In fact, it may inspire new options by allowing the traveler to see attractive options that would probably not be revealed by a hyperphemeral search to find one point-to-point schedule.

Short-term vs. long term absorption

Hyperphemera is great for delivering just-in-time marketing messages. Print and broadcasting have no real effective way of bringing targeted offers to receptive audiences in the same way that the now familiar “Customers who bought this also bought…” boxes, and dynamically generated coupons for merchandise related to what the customer is browsing. This type of dissemination is ideal for this sort of call-to-action message.

At the same time, these very advantages that build on the immediacy of the web are great disadvantages when it comes to delivering more substantive advertising and identity information. Several months ago, there was a pharmaceutical company whose medication ads would pop up in my browser a dozen times a day or more. I have no recollection of the name of the drug or the disorder for which it is indicated. In contrast, I can tell you almost everything there is to know about a particular diabetes medication, not because I’m on the market for it, but because I see their advertisement on the train two or three times a week. The fact that I’m not the target audience for this drug actually indicates an even greater advantage, because we ourselves change the little demographic boxes in which marketers place us fairly often. If, G-d forbid, I develop diabetes three years from now, it’s a safe bet I’ll ask my doctor, “what about that drug?”

A lot of information has been and should continue to be disseminated in both ephemeral and hyperphemeral formats. The key question is how we choose to emphasize one over the other, and what the longer range impact on those decisions will be for information browsers, customers, and even our posterity. Hyperphermera holds tremendous advantages for delivering critical information during research, at decision junctures, and at point-of-purchase. But the lack of pervasiveness and persistence of hyperphemera should lead us to examine these choices more carefully, hopefully resulting in smarter choices about how we disseminate information.