In this Dialogue, I would like to introduce you to Peter Hatherley of Latent Semantics. He is a word lover whose love never remained unrequited. Peter’s deep interest in words led him to develop a tool for creating semantically connected word banks.
An empathetic robot that dabbles in SEO
The tool is called CISE (Conceptual Intelligence Search Engine) and it lets you “feed” it with a word or phrase and in return it gives you sets of linguistically connected words and phrases to play, optimize your content or simply check what an algorithm can and can’t do with.
Peter’s fascination and enthusiasm for diving deep into the micro- and macro-cosmos of words are engrossing.
As much as I’m intrigued and captivated by the dance of words and data, though, from the moment I first used the tool, I have all the time been persistently highlighting that you one can’t rely on a tool to be creative. And Peter, with his characteristic patience and open-mindedness, has always, equally persistently, been telling me – of course one cannot, yet this is a fascinating shortcut.
Enjoy the Silence
I know computational models can be yet another dimension of us understanding the semantics of the world outside and inside, yet I feel there will always be something that machines and algorithms won’t be able to interpret, the way we do it. One such thing is silence.
This my intuition got me ask Peter a question about silence and how would a machine read or interpret it. Ironically, although I am not surprised, as Peter’s brain child is really intriguing, when fed with “dialogue on semantics” CISE came up with the word “silence” too.
I agree, it is the silence in-between a dialogue that is “sitting on a goldmine”.
Enjoy the silence and the words in-between it.
Let’s set this Dialogue in kind of a Zen mode from the very beginning
If the many are reduced to one, to what one is reduced?
To one unified consonance of what language actually does for us conceptually. The whole aim of conceptual intelligence is to overcome the ongoing confusion of various contextual misunderstandings by exposing what morphological context actually is.
This inclusive simplicity opens the door to a richness of meaning that allows us to understand the perspectives of others quite apart from opening the door to creativity.
Words, Love and Latin
Why do you love words?
What a great question! I love words for what they can become. Yet even more so, I love the conceptual tapestry that they can weave within the patterns of language.
Semantic fusion is based on the weaving of one or more of these contexts into your overall concept much like a metaphor does. To me it is like painting a vibrant picture much like a glorious sunrise. I love the fact that they are infinitely potent not only for ourselves but also for others.
What do you think about the possibility of Lingua Franca 2.0 :)? Could that be a breakthrough for semantics and kind of closing the circle of misunderstanding?
Here’s what’s behind this question, Peter and what I mean. Recently I stumbled upon this: Generating Cross-Linguistic Semantic Maps and I thought what if a Universal Dictionary of terms in the Western Civilisation (I know this is a bit crazy and the name is awful) could be developed from Latin, fed by Latin words and their historical development throughout ages (accrued connotations, other meanings) and all this with mapped relationships.
Etymology played a major part in my research yet the one overriding factor for me was the synchronicity within all languages not just individual ones. The original insight came from studying ancient languages.
Yet when you ultimately perceive that an ancient language can effectively interface with a modern one it becomes a discovery that can not only put aid to the confusion that we see within different languages but even more so it opens the door for confusion to be removed within our own native languages also.
Do words have an architecture?
Words have a very definite architecture but it is in quite a unique and unexpected way. From a semantic perspective they were hidden in plain sight as it were, even to qualified linguists.
Latin is a major part of that construct but no more than Greek or Aramaic. There are threads that runs through all language and in many cases a linguistic connection or fragment can be found in one and traced through the others
Levels of Understanding
Why do you think people are so skeptic when it comes to other types of understanding and knowledge?
People prefer old solutions. Disruptive technology leads us to reevaluate our perspectives and that sometimes involves breaking old habits. This is even more difficult when it comes to conceptual comprehension.
Inclusivity is absolutely vital when evaluating something new. We need this level of flexibility, at the minimum, to embrace the future of a shared conceptual comprehension.
Having an open mind is a choice. If we want to view a subject in a fresh new way we need to be willing to free ourselves from any fixed perspectives that we have on that subject.
Who, for instance, ever knew that Google would eventually embrace floating context as a rating factor like they did with the “Hummingbird” patent?
Change is inevitable yet historically many are resistant to change. This is why old school SEO techniques, such as keyword placement, remain popular at a time where they are virtually fading away beneath the fabric of semantic optimisation.
Do you think there could be understanding without a paradigm to set its parameters and how far do you think understanding can go?
Language in itself is a paradigm. I believe conceptual understanding is inherently hidden within words and sayings.
So communication without language is extremely difficult because in the end everything can be named and described.
Conceptual memory in itself appears to be intrinsically connected to words and this can vary from person to person. Yet wisdom is the ability to know how that simplicity in communication produces a far wider effect than complexity ever could.
I have the feeling that misunderstanding and miscommunication are rather a choice or a matter of protocols. What’s your take on that? Where does misunderstanding come, is it only misinterpretation?
Interesting angle. Again it comes back to context and layering that context effectively. The easiest way is by dressing your entities more effectively. A metaphor provides a vehicle of shared understanding that provides an accepted foundation where new information can be added onto our existing conceptual understanding.
By carefully considering our contextual alignment we can effectively avoid unnecessary misunderstandings. Clarification brings clarity and in some cases this is absolutely vital.
Describing a Universe of Universes
How is the dynamics of language interwoven in your tool CISE?
The algorithms look intrinsic patterns and subtle variations within every word or idiom.
Language is built out of fragments and because those fragments retain a perceptible memory of the knowledge of those fragments (otherwise known as semes) there is a dynamic result wherever these fusions occur. So, when we evaluate more complicated word fusions, it is important to retain this knowledge at the forefront of our considerations.
The dynamic nature of language allows us to be creative with these individual fragments as we see in the creative use of hashtags on G+ and elsewhere.
CISE provides concepts not keywords per se and yet the initial query is far more vital than with any other search engine. I have been thoroughly amazed by the fixed mentality of searching for keywords and keyphrases during the beta testing stage. For me it has always been about stretching the limits and that includes investigating conceptual searches at their roots.
A creative query will always get a creative result. So, it’s not like keywords in, keywords out but if you want to write conceptually you gain far more by thinking semantically.
The results are morphological in nature meaning, so it can take an existing idiom and transpose the context of it quite intuitively. So the query must be kept in mind as being the most integral part of the concept and even though we are provided with highly related concepts the opportunities to create innovative fusions is virtually unlimited
Existing idioms are effectively semantic fusions that have previously gained traction in communication. When they are transposed into a new context their meaning can morph to effectively expand our conceptual memory in the process. This morphs much like a comedian plays with context.
A good example of this phenomena is a recent article I wrote on G+ relating to semantic infusion in web design. The concept of body language was exposed in a search on web design. This became a perfect vehicle to explain how semantic optimisation works.
In a non-networked environment (if we could still find any) where is metadata stored?
In people. In our memories and especially in our collaborative fusion. Again it comes back to conceptual memory and this exists whether we have a network or not. Shared metadata is the key here and yet networking as a word has morphed to mean exactly what our digital society has become, even though it has taken on many other guises in the past.
If you fed your tool with the dictionary of untranslatables would it perform better or rather this will lead to clutter/ stack overflow?
Because it is an algorithm it has no problem analysing new information.
I have found this method particularly useful with archaic words and symbols where the meaning has been lost or is unknown. For instance, a deeper understanding of the patterns through long term exposure to them allows me to read Old English as if it was a modern language.
Yet CISE takes it even further. It can take new words and analyse them according to these intrinsic patterns. In fact I particularly like analysing new words and have added some of these to the resource due to their inherent usefulness.
Sometimes we have words without opposites so it is nice to fill in the gaps like the use of unbias as a double negative that positively explains the mentality of inclusivity.
On Metadata, Conceptually
Are there limits to describing? And what happens when metadata (data about data) exceed data?
The thing with semantic connections is that they run very deep within language and one of the major factors in developing the perfect dataset result is based on scalability. Context is the seed of understanding so by isolating context be can potentially enhance our communication from a marketing sense
The trawling side of the CISE algorithms is particularly good at isolating context produce contextually isolated Silos in levels. This is a really exciting development as it creates a drill-down mindmap of these concepts that can isolate whether a post has achieved what you intended it to or not.
This is known as sentient analysis but the major point of difference is the exposure of the major concepts suggested by any one part of it. This means comments can also be rated by considering if any pattern based value was added to that conversation.
This compact small data like solution allows quick and valuable insights into the subconscious effects of your content rather than having to scour huge lists of data to fin what you’re looking for
The CISE algorithm isolates and rates the relativity of words through their inherent patterns and these are reflected in the order of the results.
This relativity can reach to thousands of words. Yet our drill down silos keep it simple for those who are developing web structures, online courses or books
Is there an untranslatable thing (e.g. silence)?
An interesting philosophical question. Are feelings expressible? Is a scene indescribable?
We used to think images and videos were untranslatable until earlier this year but know they’ve become like words to the Googlebot these days. So the question remains why did they do this?
Sometimes we see something and say words couldn’t express how we felt in that moment but in reality they can. Silence is still describable and as always if there isn’t a word for something you could soon make one. Yet the experience in some ways transcends the words. If we perceive what we say as virtual triggers that manifest like a command at the prompt or seeds that morph into the fulness of their potential we can then use them to create silence for example.
Less is definitely more when it comes to activating change.
If language is the house of being, what is metalanguage (metadata)?
This is an extremely insightful question.
In my view, metadata is like the seeds or building blocks of conceptual understanding. Yet in many cases metadata is no more than a glorified index or keyword structure.
Because CISE automatically produces conceptually related metadata for on-page optimisation you then effectively start with the foundational elements from the outset
Schema remains a limited subset of this process all be it a very important one. Yet on page semantics is virtually unlimited and can create a metalanguage all of it’s own.
Metalanguage in essence should connect us to the wider matrix of understanding otherwise it is ineffective. The elements within that interlingual fabric then in turn provide a unique door to the wider semantic landscape.
Isn’t the only truth (in both philosophical and logical sense) the shared one?
Absolutely. Language is built for sharing that’s why the positivity and adjacent context sections within the APP are concepts we can all agree on.
For instance we might not all believe in a god per se but we can all believe in doing good. So in a way this is perfectly described as the shared part. By pooling all religion and belief as being entirely neutral we provide a path that is not polemic but inclusive. From Semantics through to semantic awareness and all the way to semantic optimisation.
Semantic Analysis
What is semantic copywriting?
For me it is conceptual copywriting that conveys deeper meaning and connectivity. Obviously this depth includes structural logic such as schema and JSON-LD. Yes, CISE provides conceptually related elements that we can place directly within our content but not only that it can provide the methodology to enhance the conceptual reality and creativity within that subject.
Semantic infusion allows us to increase that relativity quite naturally due to the lack of restrictions for available placement.
If you were to describe your tool with a metaphor what would that be?
This is a difficult question because it has such a wide specification. It is a problem solver, an encourager and even an SEO optimiser. So let’s just say my chosen metaphor for CISE would be an empathetic robot that dabbles in SEO.
How does semantic analysis cope with metaphors?
IMO it is perfectly suited to the use of metaphors especially now that Google is rating floating and adjacent context. This is even more obvious within CISE itself as the number of morphed idioms you can utilise within the professional version are quite broad.
Idioms are like mini-metaphors that automatically lend themselves to the subject at hand. So in effect CISE provides creative solutions through these metaphorical connections. Yet the name in itself is a metaphor as it took almost 12 years for it to eventually appear.
What does being semantically aware involve?
Semantic intelligence is based on the foundations of integrity, authority and inclusivity. It takes into account that others might have a quite different perspective and if they have that immediately becomes part of the ‘semantic’. I have recently expanded this concept to build music playlists based around semantic concepts called Semantic Tunes.
The expansion of meaning is never ending and I see CISE as a means of exploring these factors from a point of connectivity, right from the very start. For me, CISE is like a semantic portal that takes you directly from any one of the conceptual results to a fully fledged Google search result where you can explore the subject at hand more deeply.
Quick Favourites
Peter’s favourite super hard to describe word: morphological
Peter’s favourite similarity in the function of the left brain and the right brain: the two being merged as one entity
Peter’s favourite intersection between linguistic and computational theory: natural language processing
And now, Peter, I would be grateful if you could spare several minutes to ask me a few questions.
Peter: What resources do you currently use to enhance connectivity within a subject?
Teodora: Intuition, deep diving into my mind and into the essence of things, trying to reach their root/reason.
Peter: What is your favourite language format for use in schema?
Teodora: :) If I got the question right, it’s RDF Schema (not without its best friends OWL and SKOS)
Peter: How important is onpage connectivity to a visitors engagement, in your view?
Teodora: Very :) Let me cite my favourite poet: Sir Tim Berners-Lee: Meaning is web-like. :)
Peter: Why do you love semantics?
Teodora: Because it feels like metaphysics in action. A desperate attempt to describe the undescribable. A dance with the immaterial. A dance to the rhythm of silence.
Or to make things lighter: Shiva Fighting a Fly…