Issues of information disorder online could be reduced to the matter of information quality: not just about distinguishing truth from falsehood but highlighting legitimate and credible information. In the online information ecosystem, ‘trustworthiness’ is key, an emphasis that trending content online should have a signal of the level of ‘trust’ we can place on such content. Some major social media platforms have recognized the importance of signals of trust and credibility, for low-quality content to be de-prioritized and high-quality amplified. According to Nick Clegg, VP of Global Affairs and Communications at Facebook:
Facebook is also in the relatively early stages of exploring whether and how to rank some important categories of content differently — like news, politics, or health — in order to make it easier to find posts that are valuable and informative. [1].
Information disorder – misinformation, disinformation and malinformation – is as old as time. What is distinctive about the present day propagation is the speed, reach and virality, amplified by online social media. For example, during the 2016 US presidential elections an estimated 126 million Americans were exposed to online content, containing misinformation, that was sponsored by a foreign country. Such reach would not have been possible without social media, certainly not at such speed. Researchers, policy makers and analysts have revealed social media platforms culpable for amplifying misinformation. The obvious solution would be for online platforms to de-prioritize misinformation (add friction against virality) and elevate genuine information (amplify for virality).
Avram Turing’s Startup, called ContentQual®, deploys digital technology for online misinformation prevention. It is a consumer-facing/content-focused online platform/app/Web browser extension. It curates trending content found on the Web, helps to amplify high quality content circulating online and add friction to low quality/misinformation (slow down/reduce the sharing).
ContentQual uses a Wittgensteinian analytical check system: a thesis about information quality focused on ‘meaning’ [2]. All content viewed through this quality check device is recorded as either analytical or non-analytical (logical or illogical). Statements with no meaning (no semantic content) possess no information. This analytical check is restricted to factual language/declarative sentences.
What really is ‘information’? It is important to be clear about a definition of information. Information scientists have adopted a general definition of information (GDI). According to the GDI, information is data that is well-formed and meaningful [3]. Misinformation and disinformation are not genuine information because they are false, although they may have semantic content. Semantic factual content is what distinguishes authentic information from false information [3] [4] (in declarative language).
Information quality is an issue primarily about meaning, not so much about grammar. A sentence can be grammatically correct but have no meaning. Language functions properly only when it expresses meaning. In the examples below, all sentences are grammatical but only the sentence (3) contains authentic information. Sentences (1) and (2) are meaningless and (4) is a hypothesis:
(1) That man, the human next door, he is actually a reptile.
(2) He murdered that innocent man, but because he didn’t really kill him, the victim is dead.
(3) That tomato you are holding in your hand is a fruit.
(4) The sun will rise tomorrow.
Unless further evidence is given (or the presuppositions made clear [5]) to support the claim made in (1), an organism cannot simultaneously be human and a reptile. It is possible that the writer of (1) has some secret knowledge about humans and reptiles and the belief expressed could be justified. Knowledge can only convey information when it is made concrete and articulated through the mind [6] If there is tacit knowledge hidden in the mind of the writer of (1), it is not expressed coherently: no information is communicated. If belief in (1) could be justified, that would not make the statement necessarily true [7]. Sentence (2) could be further clarified but as it stands is meaningless. The claim in (3) may appear strange to many people (a tomato is generally thought of as a vegetable), but, even without support (clarification in the sentence), the statement is authentic because it is scientifically correct. Stating that a tomato is a fruit is different from saying that the sun will rise tomorrow. The claim in (4) is not fact but a hypothesis because what is asserted is really speculation based on historical antecedent. Information can be regarded as the product from data processed through a filter of facts, logic and semantics (see figure below):
Figure: A device/filter or information quality check separating information from data.
———–
Notes
[1] N. Clegg (2021). You and the Algorithm: It Takes Two to Tango. https://about.fb.com/news/2021/03/you-and-the-algorithm-it-takes-two-to-tango/ [2] U. Omoregie (2021), Our Approach to Misinformation Analysis: Logical Atomism. https://avramturing.com/blog/our-approach-to-misinformation-analysis-logcial-atomism/ [3] Floridi L. Semantic conceptions of information. The Stanford Encyclopedia of Philosophy, https://plato.stanford.edu/entries/information-semantic/ [4] Levitin DJ. A Field Guide to Lies. Toronto: Penguin, 2020, p. xv. [5] Yablo S 24.251 Introduction to philosophy of language. https://ocw.mit.edu. [6] Stenmark D. Information vs. knowledge: the role of intranets in knowledge management. Proceedings of the 35th Hawaii International Conference on Systems Sciences, https://www.computer.org/csdl/pds/api/csdl/proceedings/download-article/12OmNymjN0R/pdf [7] Gettier E. Is justified true belief knowledge? Analysis 1966; 23: 121-3.