The query "to our surprise" currently shows around 219 000 results. The query "not surprising" has about 1 620 000 results. That's about 7 times as many.
I have no idea if this is surprising or not.
Somewhat more surprising is that the query "italian mad scientist" has 4 results in Google Scholar. Those are all due to the SCIGen fake scientific paper generator, which sometimes outputs this phrase: https://pdos.csail.mit.edu/archive/scigen/
I would never write "to our surprise" in a paper. Rather, it would be "Surprisingly, ..." which fits the pattern of "Thus, ...", "Furthermore, ...", "In contrast, ..." etc, which is a nice and compact way of making the structure of an argument clearer.
Surprisingly, when I do the same searches as you, I don't get anywhere near the same numbers.
“To our surprise” and “surprisingly” do not mean the same thing. “Surprisingly” means that something is objectively surprising, or that a “reasonable man” would be surprised by it. “Surprisingly, hot water freezes faster than cold water.” That is pretty objectively surprising.
“To our surprise” means only that you were surprised by it. It’s more humble because it allows for a reader to be smarter/wiser and not be surprised. “I thought Frodo was going to die at the end of the Return of the King, but to my surprise he did not.” “Surprisingly” wouldn’t be correct here since not everyone would have made that prediction.
Good distinction. As a scientist, 'to our surprise' is exceptionally surprising!
We are already domain experts on the research edge for whatever topic, and writing to our peers. What we know is already not obvious to non-experts. For something to be surprising to the domain experts too, that's where the novelty points start, and why we are taking time to write/read. So, clever!
I’m struggling with the concept of objective surprise. Surprise is a subjective experience usually brought about by events differing from expectation, but the experience is entirely subjective. So my guess is that you mean the fact of results differing from expectation.
There can be a difference between results and consensus expectation, but there can also be a difference between results and personal expectation (e.g., results do not support hypothesis). In the absence of explicitly calling out who holds the expectation, there is either context that makes it clear (e.g., the audience is experts in the field and the expectations are widely held) or the use is referring to the author’s own surprise.
Simply claiming they are not equivalent is not correct. They are equivalent in the absence of context. It is only when context is considered (and just being a scholarly article isn’t necessarily enough context) that they might be considered non-equivalent.
Scientific experiments also try to prove surprising results.
"We believed this surprising thing might be true. To our surprise, nothing unusual happened."
What's generally surprising is not the same as what is surprising given the researchers' prior expectations. (What is surprising also changes over time, as a result of those experiments!)
When someone writes, “Surprisingly, …,” they are expressing their own surprise. It is equivalent to “To our surprise…,” except it is more succinct. Absent other context, there’s no implication that anyone else is surprised or should be surprised.
The “painting” is taking place either in additional context or in assumptions the reader is bringing into the interpretation. It is not present in the bare text.
It’s an ungenerous interpretation to suggest that people using “Surprisingly,” are immodest/imprecise/etc.. I could as easily hold the interpretation that “To our surprise…,” is the author trading my time in an attempt to get himself some goodwill from readers. However, that would be equally ungenerous.
How is "to our surprise" trading your time compared to "Surprisingly"? Is it because it contains more words?
The modesty here is about saying "We're not sure if this is generally surprising for everyone, but we were surprised". "Surprisingly" is a bit less conservative to me, its more of a "We think this is generally surprising for the field". I would personally use it only if I knew something has already surprised multiple research groups
The point isn’t how, but that the idea isn’t in the words. If I held that view, I’d be bringing in assumptions into the words that aren’t actually there in the words. This is true even if the author actually had those thoughts. The words simply don’t carry that information.
Human language is a social construct, and words carry whatever information users have decided to assign to it. Looks like quite a few language users disagree with you
Yes, you and people like you would read it as that they are expressing their own surprise, because you have a firm mental model that surprise resides only in the mind of the observer. But quite a few people also allow for a pseudo-objective meaning, as in "many/most people would find this surprising". Because the author may or may not be like you, it's hard to know which meaning the author intended, so if you want to communicate clearly it's best not to use the word "surprisingly" at all.
An author’s intent doesn’t change what an author actually said, nor does it mean we should assume it means something different. “Surprise” in any form should usually not be used, but if it is, the two phrases are still equivalent unless the context dictates otherwise or the author makes his use explicit.
>An author’s intent doesn’t change what an author actually said
This is an odd view of communication. It works for whole stories, e.g. you can decouple Heinlein from his books and claim that The Moon is a Harsh Mistress primarily says something other than Heinlein actually had in mind, artistically speaking, but it's a very odd stance for single words. Communication is two-way, it has everything to do with both the speaker and listener. Words don't have any meaning except for what the speaker and listener assign to them. There is no "what the author actually said", there is only what they intended to say (and how it was interpreted).
In the role of listener/reader, your whole job is to figure out what the author intended. Otherwise I'm not sure we agree on what reading even is.
We don’t have two-way communication or any context to understand the intention. We have “To our surprise…,” and “Surprisingly,….” These phrases in the context of writing, absent other context, have the same meaning. This isn’t really an opinion.
An author’s intention cannot be inferred without additional context. So no, the author’s intention per se does not matter in the context of this conversation. Extending that to mean it never matters isn’t at all what I said.
Whenever a speaker is talking about opinion or emotion, it is the speakers' opinion or emotion unless otherwise specified or contextualized. If I say, "We should eat pizza," the correct interpretation is that it is my opinion we should eat pizza. If I say, "In my opinion, we should eat pizza," the meaning is the same.
Surprise isn't an objective fact. It's a statement of emotion or subjective experience or whatever you'd like to call it. It is often a consequence of an objective difference between an expectation and a result. (An expectation can be subjective, but whether someone holds that expectation is either true or false.) But surprise isn't that objective difference, it's the subjective response.
So if I say, "Surprisingly, the sun rose today," absent other context it means I find it surprising that the sun rose today. It doesn't mean, "Everyone should find it surprising that the sun rose today." For that, I would need to add more context or you would need to bring in some assumptions that aren't supported by my statement. So, this is equivalent to "To my surprise, the sun rose today."
It also dates the paper - even if it was generally surprising 10 years ago, it won’t be today. Although I doubt papers with this language are standing the test of time anyways.
Ah yes, what kind of academic would ever write three words where one have the same meaning... And make you this much closer to fitting in that darned limit.
"Scienziato pazzo italiano" fares slightly better at 16 results (that's a literal translation; I'm sure there are better ones). Amusingly, one of the results refers to "John Lithgow", which I instantly recognized from Buckaroo Banzai - one of his characters is "Dr. Emilio Lizardo"
At first, I fully agreed with you. Then I saw another comment using "an experiment was conducted" as an typical example of passive voice, and I find that to be far more passive:
"...was conducted" suggests no information about the conductor. It could be the author, and in the specific context of an academic paper it implies the author, but it could just as well be anyone else on Earth especially outside of this context.
"Surprisingly" (or "the results were a surprise," etc.), on the other hand, doesn't leave anything to the imagination. I can't think of a way this word could potentially mean that another party but the author was surprised.
Passive voice could hide the party responsible for surprising someone (active "Alice surprised Bob" becomes passive "Bob was surprised"), but in the original scenario, the party causing the surprise (the subject, which could be hidden by passive voice), is the experiment. The surprised party is the direct object, not the subject... passive voice is not a term to describe hiding the direct object; that's transitive versus intransitive.
With all due respect, if you're not familiar with the concept of voice in linguistics (which has a particular technical definition, and makes no sense to apply to a purely adverbial phrase like "surprisingly" or "to our surprise"), maybe you should read up on the concept before trying to determine the difference between active vs. passive.
The term distinguishes between sentences where the semantic agent is the syntactic subject, and those where it is the object.
For example: "Bob threw a ball at Suzie" (active, Bob is the syntactic subject) vs. "A ball was thrown at Suzie by Bob" (passive, Bob is the object of a preposition and "a ball" is the subject).
I edited the tail end of my previous comment to touch on this, which is to say that omitting the direct object typically involves using intransitive instead of transitive. For example, the intransitive "today I ate" instead of the transitive "today I ate pie" leaves you guessing what I ate. The verb "surprise" does seem to have an intransitive syntax per https://www.merriam-webster.com/dictionary/surprise which gives a similar sentiment as "surprisingly."
I'm not sure I agree that a distinction needs to be made between the two, certainly not on lines of agency. The "our" in "to our surprise" may or may not correspond to an agent, patient, etc., in the modified clause.
To an agent: "To our surprise, we found them. To our surprise, group A was found by us first."
To a patient: "To our surprise, they found us. To our surprise, we were found by group A."
To neither: "To our surprise, group A found group B. To our surprise, group B was found by group A."
I suppose "to our surprise" is explicit about whose expectations weren't met in a way that "surprisingly" is not. But in a first-person narrative, I imagine most readers would understand "surprisingly" to mean "to my/our surprise."
This suggestion is precisely comparable to the passive voice.
There is an intended agent experiencing surprise, and we might under some circumstances agree that we know who the agent is (just as we expect that "the experiment was conducted" by the author as researcher).
But, technically, the agent is obscured and not written, so we can't be completely certain about the writer's intent. Maybe Rosalind Franklin conducted the experiment: we are only led to infer that the author was responsible.
Personally, I generally understand "Surprisingly," to mean "An attentive reader should now be surprised that...".
There's an expectation of general surprise relative to an earlier claim in the text; the writer assumes the reader will be surprised (whether or not the writer was truthfully surprised). I find myself annoyed by this style, whenever I am unsurprised.
Comparable to one use case of the passive voice, maybe.
What do you mean by "intended agent"? "Agent" has a common definition in linguistics, and an agent is only an agent in the context of a verb, not an overarching narrative. The same referent can be an agent in one sentence but not the next: "We hid. To our surprise we were found by group A." "Group A" is the only agent in the second sentence.
There may be other definitions of "agent" in other contexts, but we're talking about grammatical voice here.
The passive voice is quite simple: the grammatical subject is filled by the patient or theme of a clause, not e.g. the agent. There are many use cases for the passive voice apart from obfuscation, and obfuscation is hardly a necessary result.
I might call "surprisingly" in your interpretation a weasel word in the broad sense. The inference is that the author wants the reader to take up an attitude but is not being forthright about it.
Of course some passive clauses may be weaselly in their own right, but the passive voice is not weaselly by definition.
I think you're right; it's my impression that passive voice has been the traditional style for scientific papers. However, the winds of change have been blowing for the last 15 or 20 years. Editors of most scientific journals now recommend using active voice wherever possible.
I frequently fight with my collaborators over this. In a nutshell I think the passive voice style of science writing is unclear, and dishonest in that it pretends the researcher was not involved in the research.
My high school science classes taught the "an experiment was conducted..." passive style of lab report writing as mandatory for science writing. That, and the idea of long, clause ridden, sentences. It wasn't until 20 years later when I went into industry that I realized how bad "traditional" scientific writing is. Clear, short sentences. Active voice. No flowery or euphemistic language. All make things so much more readable. But I think many of us were raised to associate simple writing with simple ideas, and taught we had to write things that only smart peopl could understand. I'm happy to see science writing trying to be more readable.
I find that clear and simple writing correlates with high quality work on the technical level. Especially at the top of the respective field.
For example in AI if you read the papers coming from the best labs the writing tends to be on the point, short sentences, almost informal-sounding, actually someone telling you what and why they did and what they found. Not merely conforming to arbitrary rules set up by bitter mediocre "scholars".
But yes, people who are afraid they have nothing to say try to hide it behind long sentences with incomprehensible nested structures and overjargonized vocabulary. Plus equations that are designed to impress rather than enlighten.
> Plus equations that are designed to impress rather than enlighten.
Interesting, I don't think I've seen any discussion before about clear, unpretentious presentation of math. I'll have to consciously evaluate this better as I read papers.
Personally I think mathematical clarity is tougher to evaluate because I would not be confident enough to always know if it's the author or just me.
That sounds like a good indication that the math you are thinking of was written to impress, rather than enlighten, you as a reader. I've seen far too much of that stuff in computer science papers, where dense math seems to be used defensively, rather than to bring clarity.
Yes, I hate it when they handwave over the tough core issues but then flesh out in detailed notation things that are unambiguous even in words, just to be able to define some Greek letter notations, which are then hardly ever referred to again in later parts of the paper.
Unnecessary equations often mean "this is totally a serious, rigorous paper, we do admirable and hard technical work here and deserve respect!".
IMO there's a good reason for passive voice in scientific literature: It's (literally) less subjective and creates some distance between the reader and the matter of investigation, which I personally find very useful. Besides, it's also often unclear who personal pronouns refer to in the first place. I have stopped counting the times I've seen single-author papers with phrases like
> We investigated X
(Who's we?)
and
> From equation (1.2.3) we see that,
(Who is "we"? The author? The author and other readers who are smarter than me? It cannot possibly be "the author and I" because I don't "see that" at all.)
It's just a stylistic tradition in academic papers, similar to the royal "we". It's just a shorthand for "an author, or a group of authors, or the whole team/organization behind the research presented in this paper". You may think it's weird, but all traditions are weird when you think about it too much.
Also, your second "we" isn't even that, I think that's just good old-fashioned generic pronoun (because English normally doesn't allow a sentence without a subject). Basically the same as in sentences like:
> In Elbonia you are not allowed to drive after dark.
> Also, your second "we" isn't even that, I think that's just good old-fashioned generic pronoun (because English normally doesn't allow a sentence without a subject).
I don’t think I’ve ever seen “we” used as an indefinite pronoun. It’s usually “one”, “you”, or “they”, depending on formality and context.
In this context, it seems obvious to me that it stands for the author(s) and the reader, in a (maybe misguided) attempt to reduce the distance between them and involve the reader in the narrative.
> in a (maybe misguided) attempt to reduce the distance between them and involve the reader in the narrative
Thank you for putting it so succinctly, this is exactly what I meant to say: When I'm reading a scientific paper, I don't want to be involved in the narrative. Don't speak to me; let the facts speak for themselves.
This would be easy to do in active voice. I could write "I did this, and I was wondering why that anomaly turned up, so I did yadda yadda" without ever involving a "you".
> (Who is "we"? The author? The author and other readers who are smarter than me? It cannot possibly be "the author and I" because I don't "see that" at all.)
I would tend to prefer “we can see” (or “we can deduce”, or something to that effect) rather than “we see”. In the first case, it’s a statement that it is possible for you to see it too, if you think about it for a while. As you say, the latter is a statement of fact that is often wrong and can come across as condescending.
Not everyone in science is a native speaker of English. In many languages it makes no sense to say "can see", seeing implies the ability. Similarly with "can hear". In fact, as a Hungarian native speaker it was very weird to me when my English teacher insisted I insert "can"s before hear and see in my writing.
What do you mean you "can hear" the loud music coming from the neighbor's flat? Do you in fact hear it or do you just have the ability to hear it and can also switch it off or something? It's a quirk of English that I just had to memorize.
This is a good point. I have been guilty of using "we" when I carried out the experiment entirely alone. I suppose it's a lie intended to increase credibility, but I never really thought about it before.
Actually, using the passive voice should have been more clear, because in most sentences encountered in scientific papers the patient is the topic of the sentence, not the agent.
Unfortunately, in English, if you want to make the patient the topic of the sentence you are forced to use the passive voice, because there is no marker for the direct object of a verb, and then the sentence becomes less readable, because in English the passive voice is constructed with auxiliary verbs, so it is more complex than the active voice.
These are defects of the English language, so the recommendation to avoid the passive voice is specific to English and also to a few other modern European languages, which have also lost the features that enable a free order of the words.
In an ideal language, it should be possible to use a standard order of the words to express their syntactic roles without additional words, but there should also exist some optional prepositions or postpositions marking each possible syntactic role, to enable an arbitrary order of the words when desired.
In such a language you could choose freely the topic of a sentence without being constrained because some choices are awkward, like when using the English passive voice.
Having written a 100+ pages paper in passive voice in German whose writing style my supervisor called "very elegant and readable", I'm not sure it's the construction using auxiliary verbs. German is very similar when it comes to passive constructions but for some arcane reason in English passive voice doesn't sound as good or natural. (Not that passive voice is natural in German – it isn't. But it still sounds orders of magnitude better than in English.)
Unfortunately, I do not remember right now some good titles, but there are various books about comparative linguistics, which analyze the similarities and differences between languages and the various existing ways of expressing the same content.
Most ancient languages, including most ancient Indo-European languages, like Latin or Ancient Greek, allowed a free order of the words, because there were markers for each syntactic role, like agent, patient, instrument, beneficiary and so on (i.e. the so-called cases).
Nevertheless, at the stage where the old Indo-European languages became attested in writing, they had a very serious defect. Even if the so-called case terminations of the words were originally a small set of post-positions corresponding to the syntactic roles, due to various phonetic evolutions conditioned by the adjacent sounds present in words, the original small set of markers had diverged into a very large set of word terminations with many distinct variants for each syntactic role, the so-called word declensions.
Because remembering such a large set of word terminations became too difficult, most modern European languages have abandoned the old declensions. Some languages use enough prepositions, possibly together with some remnants of the old case terminations, to allow a free order of the words.
English however, depends a lot on the standard order of the words to convey their syntactic roles and it has only limited means of expression for supporting a different word order.
While in the Indo-European languages the too irregular word declensions could not persist, there are other language families with a much more regular structure for the syntactic role markers, while still having enough markers to allow a free word order.
It is weird that the European linguists of the previous centuries considered the regular languages (the so-called agglutinative languages or isolant languages) as "primitive" while considering the "flexionar" classic Indo-European languages as "superior" and "advanced".
In fact even if the flexionar classic Indo-European languages evolved from some older agglutinative language, their irregular declensions were a serious defect and not a sign of progress.
Unfortunately, because the language changes have always been done mostly by the less educated people, without having any kind of grand plan of how to best improve the language, the necessary simplifications of the language have also been frequently accompanied by a loss of expressiveness, like in English.
Thanks! When I was pretty young, I studied a language that used word terminations (and other means) without word order. I was too young too appreciate the linguistic differences, nor did I understand why a certain word order was chosen. It would give a lot more flexibility.
I am going to lookup what you wrote about the evolution and use of active voice in English - something I always wondered about.
I was a doc reviewer in a commercial organisation, checking both technical reports and commercial writing (bids and contracts).
I recommended to authors that they should use active voice because it usually made it easier to explicitly identify the intended actors and avoid ambiguity about who should do what. E.g. a bid author should prefer "$COMPANY will do X" rather than "X will be done" because it avoids confusion over whether X is a $COMPANY or $CUSTOMER responsibility. Obviously authors could still write "X will be done by $COMPANY" but I found that they would often forget to identify the actor if they chose passive voice.
I fought so hard with that in grad school 10 years ago. A number of newer papers in my field were written in the active voice, and I dramatically preferred that approach. My supervisor, however, was adamant that everything needed to be in the passive voice. He won, naturally, because I needed him to sign off on things. I do wonder how different that conversation would be today.
Exactly. The only "right" way to choose active vs. passive is to decide what you want to emphasize.
When describing an experimental method, it literally does not matter if a Nobel laureate, fresh-faced undergrad, robot, or zombie added the reagent. The point is that it was added, and so the passive voice works fine.
However, perhaps it is important that the reader understands how carefully patients were evaluated, rather than the fact that they were evaluated at all. Then, you might choose the active voice and write "A panel of 17 experts carefully evaluated each patient" instead of "patients were evaluated by 17 experts."
Indeed. Choosing one or the other is making a choice about what should be emphasised. Any form of writing is subjective, and scientific writing is not different.
When I taught writing in grad school, I actively (pun intended) pushed students away from passive voice. Active voice is generally easier to read, and typically forces the writer to clarify more things within a sentence. There are, of course, certain instances where passive voice better suits the context.
Academic writing is needlessly dense for all kinds of reasons other than passive voice though.
Interesting that there was teaching on how to write at all. In my technical PhD in Germany we had zero instruction on writing other than feedback from the prof near deadlines. You're just supposed to pick it up by reading papers (plus the experience of having written a bachelor and master thesis where you're guided by a PhD student who may themselves not be great at writing).
When I went to uni in Sweden, there was a mandatory course in academic writing as part of CS program. The learning objectives was more than just a styleguide proofreading by supervisors though, as methodologies, data collection and paper structures was covered. Pretty much the scientific process really.
This seemed to be quite needed for students to get up to speed. I was from Finland, where this subject was brought up before uni.
For what it's worth, I remember my supervisor calling bullshit on my then dense and overly academic writing and I thank them for it. It was a process of un-learning bad patterns or preconcieved notions about the writing process.
We have seminar courses for this where master students do some lit review (mainly based around one paper) and present that work in a longer form writing.
This is also "just" learning by doing. There is some back-and-forth with PhD students who help with the structure, but there's no explicit instruction.
Oh good god :) Picking up writing from reading the average paper is likely injurious to the health of your future readers. Few papers are well-written and you're likely to unknowingly pick up bad habits from reading the average paper. If your advisor was actively training you in good writing and providing you critical feedback on sentence construction, paragraph organization, the flow of ideas from one section to the next, then sure, you wouldn't really require a class. But I doubt the average advisor has the bandwidth for such feedback, esp. in my field of biology.
In the context where I taught, most students coming into the PhD program had no training in writing whatsoever. The first round of essays that people turned in were typically very poorly written. Most folks really needed the training in my view.
I mean, you wouldn't emulate the average paper though, but the ones from the best labs in your field, the papers that get awards etc.
I think it's actually better to seek writing advice on the Internet (not from randos, but if you're cut out for a PhD you should be able to tell apart crap advice from good ones.) I'm pretty sure it's better than whatever my university could offer as a course.
And just as everywhere, 90% of everything is crap. Perhaps more than 90% of papers are crap and 90% of academics have nothing to say and write terribly to hide it.
Fair enough. It does sound like you fall/fell on the more advanced side of the grad student curve though. +1 on 90% of papers having nothing much to say.
> typically forces the writer to clarify more things within a sentence.
Hmmm I don't think that's the case at all – if at all, active voice is less clear and "sloppy". (See the examples I mentioned here[0].) But maybe we interpret the term "clarify" in different ways?
Right, I would say that's just sloppy writing in general and not sloppy because of the use of active voice (and certainly an example where passive voice is better).
You're right in terms of what I meant by "clarify", which was more about writing that looking less "sloppy". The examples I think about aren't ones where the investigators are referring to themselves by "We" but where you're writing a paper about molecules interacting with each other. So something like "A activates B, which in turn inactivates C" reads more smoothly than any of its passive-voice counterparts.
> So something like "A activates B, which in turn inactivates C" reads more smoothly than any of its passive-voice counterparts.
Yes, because there is a clear subject and the fact that one thing activates another is important (at least in scientific papers, causality is kind of the point). Using the passive voice in this case is clumsy and awkward. On the other hand, there are legitimate cases for the passive voice when the subject is unclear/unknown/unimportant. Forcing the active voice then results in the overuse of meaningless pronouns or vague words just because there needs to be a subject. Good writing is using the right construct, which depends on context.
I wouldn't use either. People tend to overuse adverbs, in my opinion, perhaps thinking that by adding them it strengthens the statement in some way. To me, though, it just sounds weak and forced.
My first attempt at generated English sentences was while avoiding a middle school poetry assignment. I was just basically selecting nouns and verbs from quickly hand made list of words. It produced: "the duck hit the bus with bike". I was a bit excited it meant something, I was totally disheartened that in addition to 5th grade grammar I was going to need to teach my APPLE II a heck of a lot about what concepts make sense together...
Not really. Reproducing research can create surprising results (if it doesn't match what you're trying to reproduce), and novel research can be unsurprising (if practice matches with the theory).
OT: My pet search on HN is "Disclosure I work" [1] (or "Disclaimer I work" [2]). You'll see a few company names pop up much more often than others. I've been wondering why.
Some companies encourage you to to do this when you're discussing the company online. Even if you work on something totally unrelated at a company, it's a bad look if you sing the praises of a product and someone digs through your comment history and replies "Hang on, you work for the company that makes that product". Better to be upfront about any biases or caveats
You'll notice a great deal of those are from jefftk, who discloses who he works for in almost every comment.
But in general, I think it's just that certain tech companies are big -- so they get discussed a lot on HN, and also employ a lot of HN users, thus requiring more disclosures.
Just so we're clear, this is not actually "required". It's a combination of bragging rights, post credibility boost and (rarely) actual, official, authorized communication from the company.
That's rather cynical. If an employer says I'm fine to talk about the company in public forums, but they only request I give a disclosure, then that's why I give a disclosure. Not some pointless brag. You can argue that legally or ethically I'm not required, but it seems a minor thing to do.
Also, if you _don't_ disclose, there's a non-zero chance someone goes through your post history and finds you work at the company of interest, and what follows could be an accusation of shilling and attempting to deceive people by not making the connection apparent. Just as with a journalist, putting your biases upfront allows you and the reader to have a chance at a trustworthy dialogue.
My pet peeve is people writing "disclaimer" when they really mean "disclosure".
A disclaimer denies responsibility. So if you're including a disclaimer that you work for Google, you're essentially saying that you take no responsibility for the consequences of your comment because of your employer.
Charitably, I think the idea is something along the lines of: "Disclaimer: I take no responsibility for you mistakenly assuming that this comment is a independent review by someone unaffiliated and impartial.". (Though practically, it's probably just sloppy phrasing in a context where that sloppiness isn't much of a problem.)
I discovered the idea for this search a few days ago and found it intriguing. It is both broad as it is not restricted to a particular area of research and specific (for unexpected results) at the same time and seems to be directly probing the edge of our knowledge (At least I would call it that when researchers are being surprised by something in their fields).
There is also the scientific article search of the Internet Archive if you are looking for another one to search through:
Whenever I come across a typo in something I've published or in a paper I'm reading I'll do a google scholar search with quotes and see how many other people have made it. E.g. "expensively studied", https://scholar.google.com/scholar?q=%22expensively+studied%...
That reminds me - when I was younger I had trouble remembering how to spell definitely vs defiantly and would often get it wrong on my papers. So, I spent some time trying to think of a situation where it would not matter. I came up with the following: "My dad said I can't go to the party to night, but I'm [definitely/defiantly] going anyway.".
This is quite brilliant. Essentially querying certain power phrases in Google Scholar to see trends across disciplinary fields. I haven’t thought of doing this until now. Thanks for sharing.
This search is amusing because among the first few results is one referencing another:
> Although the pristine myth has been thoroughly debunked, too many biodiversity researchers fail to incorporate historical ecology into their analyses.
> Overall, we reject the notion that “the pristine myth has been thoroughly debunked” by archeological evidence, and suggest that the environmental impacts of historical peoples occurred along gradients[...]
Are we making fun of it? Indicating something as surprising is valuable because something that is surprising is something that has higher informative value. It seems to be a valuable and succinct indicator of a relevant point in the paper.
In one of the top 10 results, Wu and Liu say in 2004 that:
> To our surprise, except from those applications in
> graphics itself, GPU finds its applications in general
> purpose computations in other fields, and it comes up as a
> hot topic for research in recent 2 - 3 years.
I don't see how this was very surprising even in 2004. A GPU accelerates regularly-parallel computation where bandwidth is more important than latency.
On the other hand, you had to think in terms of "shaders", which is not very intuitive, and tooling was barely existent, so maybe it was a bit surprising.
"Sci Hub" is also popular in the science community. This is a shadow library that "collects" articles from other publishers websites and makes them available free of charge. Maybe the most important feature though is that it has a single search field and it does not matter where the article was initially published at.
> Dragert, K. , & Zehr, E. P. (20 1 2). High-intensity unilateral dorsiflexor resistance training results
in bilateral neuromuscular plasticity after stroke. Experimental Brian Research, 225, 93- 1 04.
> Attwell, D., & Laughlin, S. B. (2001). An energy budget for signaling in the grey matter of the
brian. Journal of cerebral blood flow metabolism, 21(10), 1133-1145
Just ask GPT3 how many papers resulted in proving the negation of the hypothesis, and how many papers resulted in reasonably proving the hypothesis. And maybe inconclusive results for sake of completeness. It would be interesting to see the numbers, especially in the field of psychology.
I am extremely pleasantly surprised with the results for "source codes". I thought it'd be full of people making the grammatical error of using "source code" in the plural.
It is happening so often these days that I get more and more frustrated by it...
This is typically the kind of mistake a non native English speaker could make and maybe indicative of a more international community.
At least these are "codes" not "cods", as I've seen written in a project I worked with. Real mistake by a French programmer, not a typo. Reasoning: programme->program so code->cod. In French there is no difference in pronunciation, and spell check didn't catch it. This one is obvious and funny, but you can get some weird English when you apply foreign rules to it.
We're engineers (or academics), pedantry is occasionally a hill we die on.
Trying to be precise in language allows you to communicate an idea the most effective way, imo, and at the end of the day that's you're goal. At least it should be in technical writing. If the reader has to pause when they see "codes" instead of "code" because it's just a bit off then that is a tiny failing in your communication.
Definitely not the end of the world, definitely not something to be an asshole about, but it is something I would maybe politely correct if I knew the person well enough.
Also don't take this for me saying I'm perfect in my written communication. I still don't know how to use commas and I've had them explained to me from all grammatical standpoints multiple times. Trying is all you can do.
Comp sci vs the natural sciences have an interesting split on this (though blurring more recently). Computer science tends to use "code" as a collective noun. Traditional usage in the natural sciences and parts of engineering is that "a code" (singular) is a single routine or piece of software to carry out some kind of numerical calculation or simulation. Then "codes" is just the normal plural when talking about more than one of them. So you find papers talking about things like "a new code for simulating preheating" [1] or "benchmarking simulation codes" [2].
That's very interesting, I've limited exposure to programming in the natural sciences so I've only really experienced the collective noun usage.
I had thought "a code" and then "codes" may have come down from the punch card era of programming where as you said it would be more of single routine for a specific computation which I guess then would have filtered down into scientific computing as programming evolved through fortran and others and now it seems it's all merging back together.
While the lines are often blurred, emerging idioms vs language errors are not simply a case of wanting kids to get off the lawn. It is about making a distinction between useful and harmful semantics in communication.
I would argue this is similar to the difference between accent vs pronunciation. People often conflate the two as similar. But you having a different accent makes your speech beautiful. Whereas you having a different pronunciation makes your speech unintelligible.
I have no idea if this is surprising or not.
Somewhat more surprising is that the query "italian mad scientist" has 4 results in Google Scholar. Those are all due to the SCIGen fake scientific paper generator, which sometimes outputs this phrase: https://pdos.csail.mit.edu/archive/scigen/