Wednesday, December 28, 2022

2022 in Review: A year of SearchResearch

 Once again we're at the close of the year... 

Sunset over the Pacific with a tiny green flash

... and I'm trying to summarize all that happened in SRS as a way of looking back over the year.  It's a bit of a complicated story--we certainly covered a number of topics (see the list below), and touched on a number of SearchResearch topics, methods, and means.  We talked about whether a parakeet is really a kind of parrot, where the oldest solar observatory is in the Americas (A: Peru), and how to read in language and scripts that you don't recognize.  

This was also the year where I taught the Human-Computer Interaction & AI class at Stanford (with Peter Norvig) which was a tremendous amount of fun, but I'd forgotten how much time standing up a full class takes, even if it's in a domain that you know well.  

Despite the class and a few trips here and there, we managed to keep SRS rolling along.  We had 55 posts this year, just over 1 per week and just a few less than last year. 

I hope you found these entertaining and educational--that's certainly my intent in writing them.  I want to encourage your curiosity, and give you some of the tools I've found that help me with my curiosity addiction.  

I'm looking forward to next year's crop of SRS Challenges.  Where in the world will we go next?  

If you would, leave a comment about what you found most interesting in SRS during 2022, and if you have ideas, what would you like us to cover in the year ahead?  The comment line is always open. 

As always, 

Search on!  

SearchResearch Challenge (1/5/21): A new year, a n...Carolina parakeets... or parrot?
Image identification is great, when it works.Google Lens caution
Answer: A new year, a new Challenge about parakeets!
How to find downloadable books in Google BooksHow to find open access books
SearchResearch Challenge (1/19/21): Time and tides...Are tides different in different places?
Answer: Time and tides in different places?
How to find anything #4 (part 2/3): News and Late ...How to find current news
SearchResearch Challenge (2/2/22): Search in a wo...Dealing with constantly changing names
Answer: Search in a world of changing names?
How to find anything #5: (part 3/3) Assessing Cre...How assess credibility of sources
SearchResearch Challenge (2/16/22): How can I sear...Searching for sounds
Answer: How can I search over audio?
SearchResearch Challenge (3/2/22): What are some g...How to find near real-time satellite images
Answer: What are some good (almost) real-time sat...
SearchResearch Challenge (3/16/22): Finding the c...Finding special connections and terms (krumholz)
Answer: Finding the connections?
SearchResearch Challenge (3/30/22): Where is the o...Oldest solar observatory in the Americas (Chankillo)
Answer: Where is the oldest solar observatory in t...
SearchResearch Challenge (4/13/22): Why water the...Why would you water Astroturf?
Answer: Why water the astroturf?
SearchResearch Challenge (5/11/22): Why... in New ...Symbols of New Orleans
Superb example of SearchResearch... in AlgeriaA great SRS tale from Algeria--tracking down mysterious holes
Answer: Why... in New Orleans?
SearchResearch Challenge (5/25/22): Finding origin...Using patents to find dates of old machines (apple parer and stapleless stapler)
Answer: Finding original patents?
SearchResearch Challenge (6/8/22): Why do gnats D...Gnats swarm, but why?
Answer: Why do gnats DO that?
SearchResearch Challenge (6/22/22): Why is there a...Elephant statues in Wisconsin
Where's Dan? A slight SRS delay this week...
Answer: Why is there an elephant statue in this Wi...
SearchResearch Challenge (7/13/22): What is this ...Identifying abandoned oil rigs in the PA woods
SearchResearch Bonus Challenge (July 20, 2022): Wh...US city population density
Answer: What's a large US city with very low popu...
Answer: What's this rusty thing I found in the woods?
SearchResearch Challenge (8/17/22): Horses are nat...Where did horses come from?
Answer: Horses are native to... where?
How to find anything #7: How to Find News and Late...More about how to find current news
SearchResearch Challenge (8/31/22): How can you fi...Finding those hard-to-find terms
Answer: How can you find answers to those mysterio...
Extra: Napa Soda Springs to return to life?Update on Napa Springs
SearchResearch Challenge (9/14/22): Can you find t...Finding Starbuck and Queequeg
Hint 1: Can you find the characters from Moby Dick...
Hint 2: Can you find characters from Moby Dick in ...
Answer 1: Can you find characters from Moby Dick i...
Answer 2: Can you find characters from Moby Dick i...
SearchResearch Challenge (10/26/22): A missing bui...Exploring the hole in Conordices Park
Answer: A missing building in the park?
SearchResearch Challenge (11/9/22): Questions abou...Bats: why sleep upside down?
Answer: Questions about bats--How many? Why do th...
More surprising insights about bat predation
SearchResearch Challenge (11/23/22): How to read o...How to read text in languages you don't recognize
Answer: How to read other scripts and languages?
AI writing systems, a poetic prompt, and The CyberiadThoughts about AI-generated text
SearchResearch Challenge (12/14/22): Does animal c...Does bird coloration change by latitude?
Answer: Does animal color and weight change by lat...

Wednesday, December 21, 2022

Answer: Does animal color and weight change by latitude?

 Colorful tropical birds... 

Keel-billed Toucan in Belize, Arctic tern in Alaska.  
P/C William Walker © 2016, 2010 see:

Lead us to a mystery--don't tropical birds seem much more colorful than the birds that live closer to the poles? 

Likewise, isn't there a tendency for animals to be different sizes if they're in the north vs. equatorial places?  

Even more generally, are animals whose normal ranges are near the poles generally bigger than those whose normal ranges are more equatorial?  

Or am I just hallucinating?    

Last week's Challenge was to test these two hypotheses... 

1.  Is it true that birds have more brilliant colors near the equator, and are less colorful as you go farther north?  If so, why would that be?  

I began this investigation by posing this query: 

     [ are equatorial birds more colorful than polar birds ] 

which took me to an article in the British science journal NewScientist with the helpful title: Songbirds are more colourful the closer they live to the equator.  The subtitle gives us a hint: "Computer analysis has shown that 19th-century naturalists including Charles Darwin were right: birds near the equator are more colourful."  

Game over?  Maybe, but I always recommend digging a little deeper to understand why. The idea that tropical birds are more colorful was first introduced by 19th-century naturalists such as Charles Darwin and Alexander von Humboldt (who we have discussed before in SRS--see Leaping Eels, and the Great Man Theory). 

Until recently, however, it has been hard to prove this hypothesis due to difficulty in quantifying the colors of birds.  Just looking at a few birds and observing "yeah, tropical toucans are colorful, but my local northern sparrows are drab" doesn't cut it.  What we want is some kind of broad-based analysis. We are, after all, trying to see if "birds (generally speaking) have more brilliant colors near the equator."  

Luckily, Chris Cooney (U. Sheffield) tested the idea on songbirds (specifically, the Passeriformes, which are about 60 per cent of all bird species). 

Now, equipped with more advanced color analysis methods, Cooney's group found that, yes indeed, songbirds that live in equatorial climates ARE more colorful than their temperate climate cousins.  

What's more, birds in forests are more colorful than those that don't live in forests. 

You should also know that the journal New Scientist is a kind of digest of science news.  They always provide a link to the original paper that they're summarizing, which in this case leads to Cooney et al.'s paper Latitudinal gradients in avian colourfulness, published in Nature Ecology and Evolution.  In this paper you can read all of the details of HOW they measured color (and why it's harder than you think), along with the killer plot of color by latitude: 

Bird colorfulness by lattitude. Redder is "more diversity in colors" while bluer is less diverse.  Adapted from Figure 2 in Cooney et al.' paper (see above).

Of course, I want to check another source for agreement about the color hypothesis, so I did another search in Google Scholar, using some of the terms I picked up while reading Cooney's work.  My search there was for: 

     [ bird color variation by latitude ] 

Turns out there are lots of paper on this topic!  I skimmed a bunch and found that this color-by-latitude idea has been debated widely over the years, with some papers showing true variations, while others don't.  The key difference in the findings seems to be the sample size and the method researchers used to measure color.  See Cooney's paper again--it's a hard problem! And, generally speaking, small studies (with less latitude variation) tended to NOT find color differences.  You need a big sample to see the effect. 

But as I skimmed, I noticed several references to a "rule" that described color variation by latitude.  So my next query was: 

     [ rule describing bird color change by latitude ] 

Voila!  I discovered Golger's Rule which is 
" ecogeographical rule that links animal colouration with climatic variation. This rule is named after C.W.L. Gloger who was one of the first to summarise the associations between climatic variation and animal colouration, noting in particular that birds and mammals seemed more pigmented in tropical regions." (From: Delhey, Kaspar. "Gloger’s rule." Current Biology 27.14 (2017): R689-R691.)  

Fascinating. Golger's rule is really about the total amount of pigmentation, but it extends to color pigments as well.  This color-variation-by-latitude seems to be much more than just birds, although it's pretty clear that birds DO vary by latitude, with brilliant colors being more tropically inclined.  

2.  Is it true that the closer an animal species is to the poles, the larger they are?  Again, if so, why?  

Taking the lessons I learned from the bird/color Challenge, my first query was in Google Scholar: 

     [ latitudinal variation in animal body size ] 

Which produced a wealth of insights, particularly around a rule that's similar to Golger's Rule, but for animal body size rather than pigment or color.  Bergmann's rule (1847)  was originally written as “Within species and amongst closely related species of homeothermic animals a larger size is often achieved in colder climates than in warmer ones, which is linked to the temperature budget of these animals.”  

That is, within a clade ("closely related species"), it's generally true that the closer an animal lives to the equator, the smaller it will be.   

Bergmann's rule can be summarized by this diagram (which just happens to be of a bird): 

Bergmann's rule for penguins--the more equatorial the animal within its clade,
the smaller it is.  P/C Wikimedia.


Which is a remarkable result!  But, as you'd expect, in the intervening 175 years, some exceptions have been found (including some frogs that show the opposite of Bergmann's rule and howler monkeys that are inverse Bergmann's within a limited range), and the rule has been amended for special conditions (birds, and measurements of size).  However, in general, the rule seems to hold.  For details, see this authoritative article: Geographic gradients in body size: a clarification of Bergmann's rule.   (Blackburn, Tim M., Kevin J. Gaston, and Natasha Loder. "Geographic gradients in body size: a clarification of Bergmann's rule." Diversity and distributions 5.4 (1999): 165-174.)

SearchResearch Lessons 

So, yes, birds ARE more colorful near the equator (in general) AND animals are smaller near the equator (in general).  Which partly explains the plethora of beautiful hummingbirds in the jungle.  But let's talk about what SRS insights we should take away from this Challenge. 

1. Straight-forward questions are often a great starting tactic.  The ability of search engines to understand what you mean with a question like our first query are equatorial birds more colorful than polar birds ] is improving all the time. With a good start, we can begin to dive more deeply into the real content.  

2. Double check by doing a different query to get a the same idea. In this case, my second and third queries were restatements of the original query, but using somewhat different language.  Those new terms were learned from what I was reading.  That is... 

3. Learn new terms as you read in the area. This is one of the more important SRS lessons: learn as you read.  In particular, when you see a word or phrase that's new to you, be sure to look it up--you might be able to make your next query much more focused with the language and ideas you pick up along the way.  

4. Notice ideas as you read. More than just language and terminology, you want to also notice ideas as your doing your research.  While I was searching for birds and color, I noticed the use of a "rule" to describe the latitudinal variations.   That was incredibly effective when I went to search for size variability--it turns out that biologists love to express relationships as "rules."  That's an idea I'll carry forward for the next time I have to search for something like this.  

Hope you enjoyed the Challenge!  

Search on!  

Wednesday, December 14, 2022

SearchResearch Challenge (12/14/22): Does animal color and weight change by latitude?

Colorful tropical birds... 

Keel-billed Toucan in Belize, Arctic tern in Alaska.  
P/C William Walker © 2016, 2010 see:

... are one of the glories of the avian world.  And I've noticed something odd--don't tropical birds seem much more colorful than the birds that live closer to the poles? 

When I think of jungle birds, brilliant parrots, toucans, and lorikeets come to mind.  Meanwhile, birds of the north such as Canada geese, finches, terns, and seagulls tend to be more muted.  Is this a real thing, or am I seeing something where nothing is going on?  Sure, there are brilliant red cardinals in the northern parts of the US, but on the whole... 

Similarly, isn't there a tendency for animals to be different sizes if they're in the north vs. equatorial places?  Seems to me that some animals (such as deer or foxes) seem to be bigger "up north" (say, Canada) than "down south" (say, Mexico).  There's a lot of variability, of course, but is there a general trend? 

Or am I just imagining things?  

Pondering this made me want to frame it as a Challenge for you.  Here you go... 

1.  Is it true that birds have more brilliant colors near the equator, and are less colorful as you go farther north?  If so, why would that be?  

2.  Is it true that the closer an animal species is to the poles, the larger they are?  Again, if so, why?  

The REAL question for us is how would you find out such a thing?  Remember that you'll want to not just search for positive examples of these observations, but also counter-examples.  (In truth, you'll probably want to find people who can speak with authority about such things.  Looking at one or two birds (or even a few dozen animals) over different latitudes probably won't tell you what's really going on.  

So.. how can you figure this out?  (Remember that I can be devious: one of these propositions might not be true!)  

Let us know what you discover, and HOW you found out? 

Search on!  

Tuesday, December 13, 2022

AI writing systems, a poetic prompt, and The Cyberiad

I was reading in the hammock... 

Image by Parti. Prompt: "Robot swimming in a sea of text" 

... in my parents backyard, swinging beneath a guava tree and reading scads of science fiction and old National Geographic magazines.  It was summer in Los Angeles, a perfect time for a 14-year old boy to be reading The Cyberiad by Stanisław Lem.  I knew that I wanted to do science, but at 14, I thought the science for me was going to be biology.  I'm not sure, but it's possible that reading Lem that summer, with all his depictions of robots, virtual reality, and artificial intelligence made me swerve from the biological to the computational.  

I remember that summer with delight.  And I was particularly reminded of The Cyberiad when I read a poem that was created by ChatGPT, the AI prose generator by OpenAI.  In particular, my mind flashed back to a poem that was purportedly created by a robot made by the brilliant master scientist Trurl. In this scene from the book, Trurl's arch-nemesis Klapaucius challenges his robot with a ridiculous task, one intended to show its shortcomings--it was an attempt to make the robot fail foolishly.  Here's the passage:  

“Certainly not! I didn't build a machine to solve ridiculous crossword puzzles! That's hack work, not Great Art! Just give it a topic, any topic, as difficult as you like..."

Klapaucius thought, and thought some more. Finally he nodded and said:

"Very well. Let's have a love poem, lyrical, pastoral, and expressed in the language of pure mathematics. Tensor algebra mainly, with a little topology and higher calculus, if need be. But with feeling, you understand, and in the cybernetic spirit."

"Love and tensor algebra?" Have you taken leave of your senses?" Trurl began, but stopped, for his electronic bard was already declaiming:


Come, let us hasten to a higher plane,

Where dyads tread the fairy fields of Venn,

Their indices bedecked from one to n,

Commingled in an endless Markov chain!

Come, every frustum longs to be a cone,

And every vector dreams of matrices.

Hark to the gentle gradient of the breeze:

It whispers of a more ergodic zone.

In Reimann, Hilbert or in Banach space

Let superscripts and subscripts go their ways.

Our asymptotes no longer out of phase,

We shall encounter, counting, face to face.

I'll grant thee random access to my heart,

Thou'lt tell me all the constants of thy love;

And so we two shall all love's lemmas prove,

And in bound partition never part.

For what did Cauchy know, or Christoffel,

Or Fourier, or any Boole or Euler,

Wielding their compasses, their pens and rulers,

Of thy supernal sinusoidal spell?

Cancel me not--for what then shall remain?

Abscissas, some mantissas, modules, modes,

A root or two, a torus and a node:

The inverse of my verse, a null domain.

Ellipse of bliss, converge, O lips divine!

The product of our scalars is defined!

Cyberiad draws nigh, and the skew mind

Cuts capers like a happy haversine.

I see the eigenvalue in thine eye,

I hear the tender tensor in thy sigh.

Bernoulli would have been content to die,

Had he but known such a^2 cos 2 phi!”

When I read that, I couldn't resist.  I immediately gave ChatGPT the same text, just out of the perverse pleasure of the prompt:  

ChatGPT Prompt:  Let's have a love poem, lyrical, pastoral, and expressed in the language of pure mathematics. Tensor algebra mainly, with a little topology and higher calculus, if need be. But with feeling, you understand, and in the cybernetic spirit.


of unity and harmony.

In the language of pure mathematics,

my love for you is a tensor,

strong and resilient,

able to withstand the greatest strains,

and to adapt to any shape or form.

In the realm of topology,

our love is a Möbius strip,

infinite and non-orientable,

twisting and turning,

never ending and never repeating.

And in the realm of calculus,

our love is a continuous function,

smooth and differentiable,

always changing and evolving,

but never losing its sense of direction.

Together, we form a cybernetic system,

a perfect union of mind and body,

a fusion of technology and humanity,

a symphony of logic and emotion.

In this world of ones and zeros,

we stand as a beacon of hope,

a reminder that love is the ultimate algorithm,

the ultimate solution to all that ails us,

and the ultimate expression of who we truly are.

Well, well, well.  It doesn't rhyme, and it's not as elegant in imagery or meter, but it's a poem created by an AI after following the same prompt.  

(I also have to report that when I read this poem the first time, in the hammock, I spontaneously clapped my hands in applause for the translator.  The Cyberiad was written in Polish, and yet the poem was somehow miraculously converted to coherent, smooth, silky English. You have to give that translator a standing ovation.)   

And you have to give the folks at OpenAI a lot of credit as well.  Even though I work in the field, I did not see this coming.  I have to think a lot about this in the year ahead.  What a fascinating bit of language... What will all this mean for SearchResearch?  Stay tuned... and... 

Search on!