Compress to impress

Compress to impress

Increasingly on social media, we compress our message into the most condensed format possible. Maximising the limited resource of our reader’s attention, we try to get the maximum impact for minimum investment of their attention. But what consequences come of compressed information?

Ths sntnc hs bn cmprssd by rmvng th vwls – This sentence has been compressed by removing the vowels

It’s a simple technique to get more information into less characters. However, the following sentence could have been compressed in the exact same manner to give the same code:

These sentience his bean compressed by removing thee vowels.

We know the second sentence is (most likely) incorrect, as our mind’s inbuilt error detection mechanism alerts us that the sentence does not make any sense in our English grammar system. This is an aspect of Claude Shannon’s Information Theory (SIT) playing out in our minds.

While at MIT and later Bell Labs, Shannon wrote papers that served as the foundations upon which the world we know today is built. Digital circuits, computers, the internet, robotics and artificial intelligence all find their origins in Shannon’s work. His Information Theory is expansive and for this article will serve as an opportunity to identify specific issues we encounter on the digital landscape regarding compression and decompression of information.

Compression is the process by which we take the thought form we associate with the concept “elephant” and it ends up as the word “elephant” in a conversation. We’re taking the information we could associate with an elephant and reducing it into a message we can communicate to someone else. Decompression is similarly the process by which we hear or read a word/sentence and fill in all the details of the concept using our memories and reference points.

According to SIT, the more you compress a message, the more likely it is to have errors when it is decompressed. While Shannon’s work was derived at a low level of Boolean logic (binary states), we can see applications of the theory in much higher level and complex communication systems such as human emotions and body language. On the level of human languages, an error in the message means it being misinterpreted from its original intention. The abundance of communication facilitated by the internet present us with many instances to examine the nature of these errors.

Scroll through your Facebook or Twitter feed to see information that has been compressed for distribution. Whether it’s a picture of Barack Obama entitled “Hope” or a hashtag promoting social equality (#yestoequality #blacklivesmatter), these messages have been condensed and rely on the sophistication of the viewer to decode the full message being transmitted.

In some cases, there is a strict 160-character limit but more often, the attention span of the reader defines the limit. Our messages are being modified to ensure a maximised chance of successful transmission and interpretation. A picture, said to be worth a thousand words, with a catchy hashtag or caption, can condense an essay of information into a byte-sized digestible chunk, which may lose nuance and detail but packs the biggest punch for the available attention.

Without context, #yestoequality is a platitudinous statement. However, in Ireland during 2015 for many it encapsulated a change in mindset, the accumulation of 30 years of work towards equality for members of the LGBT community. At the same time for many others, the statement represented an attack on conservative family values. The same message decompressed into two different themes depending on the frame context of the receiver.

Applied at a robotic (or digital) level, there are standardised processes and algorithms to ensure that messages are both encoded and decoded consistently. These agreed standards prevent information lost in transmission from influencing the message. For humans, there are no universally agreed standards for communication, either the sending, or receiving of messages.

On a human level, the algorithms are replaced with mechanisms of thought, developed over a lifetime and influenced by the variety of experiences in a person’s life. Family circumstances, education, religious and political affiliations, all factors which influence the way a person will reconstruct a compressed message.

A full transmission of information in a message by definition would include every detail imaginable, involved with or relating to the message. Certainly, some information compression is required to facilitate easy and efficient communication but how much compression is required? How much compression is appropriate?

The compression required is an amount such that the reader invests minimal effort to obtain useful information. The message is compacted to make it accessible to as many as possible. In this process, minor insignificant details are omitted on purpose to communicate the most relevant information.

The appropriate compression is an amount such that no significantly relevant information is omitted or lost in the communication.

Then the lines on what is deemed appropriate, what is discussed vs. what isn’t, becomes a battleground of ideologies. If I have ideas or opinions to protect from being challenged and I can protect those ideas by withholding or obscuring information, would I do so? Would someone else do so? Is being economical with the truth as bad as lying? Is it as bad to leave something out, as it is to make it up?

By reducing complex situations to palatable bites, we can do a disservice to the situation. In our quest for simple communication, we sacrifice the nuances of the issue, and with them, much of what reason and pragmatism tries to anchor onto.

The power of a condensed message lies in the manner in which you can more easily reach those who possess and understand the missing context. The weakness presents when the same message is incorrectly interpreted by those who have a different understanding of the context. When trying to work towards solutions based on common ground, are we helped by the fact that condensed information is likely to be misinterpreted? Is the information loss worth the attention gain?

At risk of losing out on people’s attention, our politicians propose simple solutions to complex problems. A simple solution with a memorable slogan is more likely to get support than a superior solution with a 25 point plan. “Stop the boats” has been the simple solution offered by Australian Governments in absence of a considered approach to addressing the concerns around migration.

In the political arena, message complexity is a risk as it offers avenues of attack for the opposition. A common political opposition tactic is to throw logical fallacy after fallacy at the argument in the hope that some simple sound bite or headline captures the public imagination and gives the opposition increased chance of getting re-elected. The Australian program has been effective at stopping the boats but has been hailed as a humanitarian disaster. The nuances in this case are matters of life and death and deserve a compassionate yet complex solution rather than a simple slogan designed to rally support pre-election.

We’re drowning in information but thirsting for knowledge.

Solutions for problems are rarely identified by people choosing wilful ignorance on a topic, so why would we think they could be best identified by people not opting into full disclosure of details?

There is a danger to forming opinions based on Heavily Condensed Information (HCI). HCI relies on having a solid context and understanding of a topic to reconstitute the message correctly. Over time, HCI inevitably reduces a person’s context and understanding, as information is omitted in the condensing. As the context and understanding reduce, the value of HCI decreases until the information becomes so context less as to make it useless. I wouldn’t suggest making any decisions based on simplified information.

Our public discourse is dominated by rhetoric designed for polarising debates resulting in binary options on choices and votes for solutions that may not be appropriate for the complexity of the problems. Take any of the big issues of 2015 and observe how the public conversation unfolds. With the Greece financial crisis, it was the left leaning Greek socialists vs. right wing Euro-centric fiscal conservatives. A complex web of interlocking fiscal and monetary, causes and effects were reduced to a Yes/No option in a referendum upon which the fate of the Euro supposedly hung.

The world of social and political congregation is best represented by a map of many colours yet the issues consistently are redrawn into black and white versions where nothing blends, and ideas only agree or oppose. As our journey in the online world charts new territory year on year, our likes and shares will steer the ship of consensus towards common solutions. It is in the long form comments where we will find the nuance and any hope of a long-term sustainable outcome. Let us not compress our news despite our Facebook*.


Takeaway – Summary Points

  • As the volume of information increases, so does the requirement to compress/condense it.
  • More compression of a message leads to more errors in the interpretation of the message.
  • Our abilities to detect and correct errors are not perfect.
  • Without appropriate context, condensed messages can be detrimental for communication.

* My apologies for the pun.  Just couldn’t help myself.

Leave a Reply