4 ms·

Nothing in there on what the involvement was, so not at all useful. Drunk driver killed self or othere? Drunk domestic partner did the killing? Drunk and killed

Nothing in there on what the involvement was, so not at all useful. Drunk driver killed self or othere? Drunk domestic partner did the killing? Drunk and killed by police for acting out? Health problems? Died because caregiver was drunk? Simply had alcohol in the system at the time of death (but also perhaps meat, sugar, nicotine, etc.)? I don't dispute that alcohol causes acute and chronic problems. But "involved" is just too vague.

BeefWellington 4y ago
If a drunk driver kills sober people, I'd argue it's correct to consider alcohol "involved" in the deaths.
- soco 4y ago
  Same if a sober driver kills drunk people, alcohol is involved - maybe even causally (they fall over the road)...
- notch656a 4y ago
  "Involved" is a nice weasel word that was invented to imply 'caused' without having to prove it.
  - happytoexplain 4y ago
    The first line of the article is "One out of every eight deaths in Americans ages 20 to 64 resulted from drinking too much alcohol, according to a U.S. population-based study." (emphasis mine) They certainly still have the burden of demonstrating that somehow, but exclusively focusing on the particular near-synonym they used in the title seems uncharitable if you have read the first line of the article.
  - bdcravens 4y ago
    That feels like a "guns don't kill people" take.
    - AlgorithmicTime 4y ago
    - notch656a 4y ago
      If I crash with a gun in my waistband, is it a "gun-involved" crash? I mean I'm sure someone has crashed because they were fiddling with a gun in their waistband.
  - xapata 4y ago
    You seem to use a very strict definition of "caused".
  - HelloMcFly 4y ago
    Seems like a more colloquial way to say something is a likely "causal contributor".
  - MichaelCollins 4y ago
    If you want to get really wormy, you can claim alcohol doesn't even cause DUI crashes; those are caused by the person who decided to drink and drive. The agency lies with the drunk person, not with the substance they consumed. But that's a meaningless distinction if you're concerned about the impact of DUI to the public. So "alcohol involved" is more accurate and gets to the meat of what we actually care about.
    - deeblering4 4y ago
      This is similar to a favorite argument from defenders of a certain US amendment
      - MichaelCollins 4y ago
        To be clear, my argument is that "alcohol involved" is not a weasel term. Trying to make the conversation about "cause" is a rhetorical trap which will turn the discussion into a quagmire about free will and particularly the agency of inanimate objects or substances. If you simply say "alcohol involved", then you avoid all that philosophical mess entirely. By analogy to the gun debate which you're alluding to, saying "gun involved" deaths would neatly circumvent the "but guns don't shoot themselves" quagmire. However in the case of the gun debate, the "cause" language is used anyway despite asking for that quagmire, because it's more emotive.
        notch656a 4y ago
        Dispensing with causation is a very convenient approach indeed. In doing so I could say "higher concentrations of melanin pigmentation is involved with higher arrest rates in the US for homicide." If questioned I will be extra certain to make it clear I don't think being black makes you commit more crime, and totally didn't mean to imply anything close to that.
- quantified 4y ago
  All the things listed there would be correctly considered "involved". But which of them are any significant fraction?
happytoexplain 4y ago
I don't understand the objection you're raising, or why the it means the information is not useful. Sure, granular data is great, but I don't see how the lack of it comes anywhere near totally invalidating the significance or usefulness of the figure for alcohol-causative deaths.
saalweachter 4y ago
> In the younger group, alcohol-attributed deaths were most frequently "other" poisonings (28%), followed by motor vehicle crashes and homicides (24% and 20%, respectively). For those ages 35 to 49 years, the leading cause was poisonings (24%), followed by alcoholic liver disease and motor vehicle crashes (18% and 12%).
smolder 4y ago
I think it's a reasonable guess that they went with the vague "involved" qualifier in order to make the number as big as possible, to bolster the core message that alcohol is dangerous and bad for people. I don't say this to detract from their message, and it's not some major deception, but I think it's noteworthy that this kind of sales and marketing treatment is given to an idea that's sufficiently well-accepted and consistent with observed reality to where it shouldn't really need the help. Personally, when I sense someone is trying to bolster their argument by cherry picking or other deceptive practice, I immediately question their motives and their message. Maybe the overall effect is that still more people get the message when it's "spiced up"? Regardless, it's unnerving to me that the tactics employed in marketing seem to be applied by default to so much public communication. I wonder how accurate my perception of this as a growing trend is, and how it might relate to the general erosion of trust networks in modern society.
philote 4y ago
It's not in the article but the study of course better defines what they looked at. They looked at deaths due to liver disease, cancer, car crashes, etc.

BeefWellington 4y ago

If a drunk driver kills sober people, I'd argue it's correct to consider alcohol "involved" in the deaths.

soco 4y ago
Same if a sober driver kills drunk people, alcohol is involved - maybe even causally (they fall over the road)...
notch656a 4y ago
"Involved" is a nice weasel word that was invented to imply 'caused' without having to prove it.
- happytoexplain 4y ago
  The first line of the article is "One out of every eight deaths in Americans ages 20 to 64 resulted from drinking too much alcohol, according to a U.S. population-based study." (emphasis mine) They certainly still have the burden of demonstrating that somehow, but exclusively focusing on the particular near-synonym they used in the title seems uncharitable if you have read the first line of the article.
- bdcravens 4y ago
  That feels like a "guns don't kill people" take.
  - AlgorithmicTime 4y ago
  - notch656a 4y ago
    If I crash with a gun in my waistband, is it a "gun-involved" crash? I mean I'm sure someone has crashed because they were fiddling with a gun in their waistband.
- xapata 4y ago
  You seem to use a very strict definition of "caused".
- HelloMcFly 4y ago
  Seems like a more colloquial way to say something is a likely "causal contributor".
- MichaelCollins 4y ago
  If you want to get really wormy, you can claim alcohol doesn't even cause DUI crashes; those are caused by the person who decided to drink and drive. The agency lies with the drunk person, not with the substance they consumed. But that's a meaningless distinction if you're concerned about the impact of DUI to the public. So "alcohol involved" is more accurate and gets to the meat of what we actually care about.
  - deeblering4 4y ago
    This is similar to a favorite argument from defenders of a certain US amendment
    - MichaelCollins 4y ago
      To be clear, my argument is that "alcohol involved" is not a weasel term. Trying to make the conversation about "cause" is a rhetorical trap which will turn the discussion into a quagmire about free will and particularly the agency of inanimate objects or substances. If you simply say "alcohol involved", then you avoid all that philosophical mess entirely. By analogy to the gun debate which you're alluding to, saying "gun involved" deaths would neatly circumvent the "but guns don't shoot themselves" quagmire. However in the case of the gun debate, the "cause" language is used anyway despite asking for that quagmire, because it's more emotive.
      - notch656a 4y ago
        Dispensing with causation is a very convenient approach indeed. In doing so I could say "higher concentrations of melanin pigmentation is involved with higher arrest rates in the US for homicide." If questioned I will be extra certain to make it clear I don't think being black makes you commit more crime, and totally didn't mean to imply anything close to that.
quantified 4y ago
All the things listed there would be correctly considered "involved". But which of them are any significant fraction?

happytoexplain 4y ago

I don't understand the objection you're raising, or why the it means the information is not useful. Sure, granular data is great, but I don't see how the lack of it comes anywhere near totally invalidating the significance or usefulness of the figure for alcohol-causative deaths.

saalweachter 4y ago

> In the younger group, alcohol-attributed deaths were most frequently "other" poisonings (28%), followed by motor vehicle crashes and homicides (24% and 20%, respectively). For those ages 35 to 49 years, the leading cause was poisonings (24%), followed by alcoholic liver disease and motor vehicle crashes (18% and 12%).

smolder 4y ago

I think it's a reasonable guess that they went with the vague "involved" qualifier in order to make the number as big as possible, to bolster the core message that alcohol is dangerous and bad for people. I don't say this to detract from their message, and it's not some major deception, but I think it's noteworthy that this kind of sales and marketing treatment is given to an idea that's sufficiently well-accepted and consistent with observed reality to where it shouldn't really need the help. Personally, when I sense someone is trying to bolster their argument by cherry picking or other deceptive practice, I immediately question their motives and their message. Maybe the overall effect is that still more people get the message when it's "spiced up"? Regardless, it's unnerving to me that the tactics employed in marketing seem to be applied by default to so much public communication. I wonder how accurate my perception of this as a growing trend is, and how it might relate to the general erosion of trust networks in modern society.

philote 4y ago

It's not in the article but the study of course better defines what they looked at. They looked at deaths due to liver disease, cancer, car crashes, etc.